Quality Control Practices for Chemistry and Immunochemistry in a Cohort of 21 Large Academic Medical Centers

Quality Control Practices for Chemistry and Immunochemistry in a Cohort of 21 Large Academic... Abstract Objectives In the United States, minimum standards for quality control (QC) are specified in federal law under the Clinical Laboratory Improvement Amendment and its revisions. Beyond meeting this required standard, laboratories have flexibility to determine their overall QC program. Methods We surveyed chemistry and immunochemistry QC procedures at 21 clinical laboratories within leading academic medical centers to assess if standardized QC practices exist for chemistry and immunochemistry testing. Results We observed significant variation and unexpected similarities in practice across laboratories, including QC frequency, cutoffs, number of levels analyzed, and other features. Conclusions This variation in practice indicates an opportunity exists to establish an evidence-based approach to QC that can be generalized across institutions. Quality control, Westgard rules, Quality control rules, Chemistry, Immunochemistry In order for a clinical test to be useful, it must be both accurate (reflect the actual concentration of the analyte in question) and precise (be reproducible). Quality control (QC) testing using measurements of QC samples (with known target analyte values) is a key component of the overall quality management process. In the United States, minimum acceptable standards for QC are specified in federal law under the Clinical Laboratory Improvement Amendment and its subsequent revisions.1 These standards are enforced through the laboratory accreditation process by the College of American Pathologists, The Joint Commission, and other organizations. The basic theoretical framework involved in the design of a laboratory QC program has been described in standard textbooks.2-4 These sources often emphasize the use of power function curves (which chart the probability of rejection vs the degree of systematic error for different QC rules) to determine the number and frequency of controls, as well as cutoff values.2 Furthermore, these power curves can also be used to calculate the probability of false negatives (accepting an out-of-control QC as valid) and false positives (rejecting QC that is actually valid), as well as to estimate the number of control measurements required to identify errors of varying magnitude.2 Some laboratories use a variation of the Westgard Multirule Chart.5,6 This chart uses a series of control rules that are employed to interpret QC data to determine if a result is either in or out of control. The rules were designed to be sensitive to both random and systematic errors.3 Typically, the rules are applied to a Levey-Jennings QC plot with control limits drawn to indicate varying degrees of deviation from the expected QC mean.6 Given that this statistically robust approach to QC has existed for several decades, one would expect that a consensus evidence-based best practice guideline for quality control in clinical chemistry (CHEM) and immunochemistry (IM) would have been developed during this time. However, to our knowledge, no national pathology/laboratory professional organization has issued detailed guidance statements regarding QC programs. Logically, one may assume that similar laboratories performing the same tests on comparable types of patients would have similar approaches to QC with minor variations due to differing volumes, instruments, and local practices. We surveyed 21 leading academic medical centers regarding their QC approaches to assess similarities and differences in QC practice across different organizations. Materials and Methods A six-question survey about QC practices Table 1 was distributed to laboratory directors at 21 large academic medical centers Table 2. These centers were chosen for their size and national reputation based on the US News & World Report 2016 to 2017 hospital top 20 honor roll.7 Because of their separate locations and laboratory management schemas, NewYork-Presbyterian Cornell and Columbia were considered separate, bringing the total to 21 instead of 20. For the purpose of this analysis, the results from individual laboratories were kept anonymous such that the QC program used at individual centers would not be identified. The questions were deliberately phrased in an open-ended fashion to collect as much data possible, with additional communication via email used to elucidate any ambiguities. The responses of the laboratories were tabulated to give a semiquantitative picture of QC practices among the participating hospitals. The instruments used were stratified by vendor company, with QC frequency stratified by number of QC levels and events per day (when “per shift” was used, three shifts/day were assumed, and startup and shutdown were considered independent time points). QC materials were divided into manufacturer (ie, provided by the same vendor that produces the testing platform) and third party (which was further substratified by vendor). QC rules were generally defined as a number of standard deviations (SDs) of difference from the mean acceptable to call a value “in control” (ie, a 2-SD rule means that a QC value is considered “out of control” if it is 2 SD or more from the accepted average). We did not delineate how these SDs were derived (eg, from testing a QC sample multiple times to determine the SD vs manufacturer-designated SD ranges). QC rules were generally in the format of “if test X is out by Y SD, we take action Z.” These rules were simplified and streamlined to give short, easily interpretable responses of uniform format (these changes were supplied back to the laboratories to confirm accuracy). Finally, the utilization of moving averages was coded as yes or no, and any narrative comments (eg, utilization for nonclinical purposes, intent on implementation in the near future) were considered separately and mentioned where relevant. Table 1 Questions Included on Surveys What instrument do you use for automated chemistry and immunochemistry?  For chemistry and immunochemistry, how do you perform QC (number of levels and frequency)?  What QC material do you use (manufacturer supplied or third party)?  How do you define your QC ranges?  What are your QC rules (eg, Westgard rules, other)?  Do you use moving averages?  What instrument do you use for automated chemistry and immunochemistry?  For chemistry and immunochemistry, how do you perform QC (number of levels and frequency)?  What QC material do you use (manufacturer supplied or third party)?  How do you define your QC ranges?  What are your QC rules (eg, Westgard rules, other)?  Do you use moving averages?  QC, quality control. View Large Table 1 Questions Included on Surveys What instrument do you use for automated chemistry and immunochemistry?  For chemistry and immunochemistry, how do you perform QC (number of levels and frequency)?  What QC material do you use (manufacturer supplied or third party)?  How do you define your QC ranges?  What are your QC rules (eg, Westgard rules, other)?  Do you use moving averages?  What instrument do you use for automated chemistry and immunochemistry?  For chemistry and immunochemistry, how do you perform QC (number of levels and frequency)?  What QC material do you use (manufacturer supplied or third party)?  How do you define your QC ranges?  What are your QC rules (eg, Westgard rules, other)?  Do you use moving averages?  QC, quality control. View Large Table 2 Hospitals Responding in Full Barnes-Jewish Hospital/Washington University, St Louis  Brigham and Women’s Hospital  Cedars-Sinai Medical Center  Duke University Hospital  Hospitals of the University of Pennsylvania–Penn Presbyterian  Houston Methodist Hospital  Johns Hopkins Hospital  Massachusetts General Hospital  Mayo Clinic  Mount Sinai Hospital  New York-Presbyterian Columbia  New York-Presbyterian Cornell  Northwestern Memorial Hospital  Ronald Reagan UCLA Medical Center  Stanford Health Care–Stanford Hospital  The Cleveland Clinic  Tisch Hospital, NYU Langone Health  University of California, San Francisco Medical Center  University of Colorado Hospital  University of Michigan Hospitals and Health Centers  UPMC Presbyterian Shadyside, Pittsburgh  Barnes-Jewish Hospital/Washington University, St Louis  Brigham and Women’s Hospital  Cedars-Sinai Medical Center  Duke University Hospital  Hospitals of the University of Pennsylvania–Penn Presbyterian  Houston Methodist Hospital  Johns Hopkins Hospital  Massachusetts General Hospital  Mayo Clinic  Mount Sinai Hospital  New York-Presbyterian Columbia  New York-Presbyterian Cornell  Northwestern Memorial Hospital  Ronald Reagan UCLA Medical Center  Stanford Health Care–Stanford Hospital  The Cleveland Clinic  Tisch Hospital, NYU Langone Health  University of California, San Francisco Medical Center  University of Colorado Hospital  University of Michigan Hospitals and Health Centers  UPMC Presbyterian Shadyside, Pittsburgh  View Large Table 2 Hospitals Responding in Full Barnes-Jewish Hospital/Washington University, St Louis  Brigham and Women’s Hospital  Cedars-Sinai Medical Center  Duke University Hospital  Hospitals of the University of Pennsylvania–Penn Presbyterian  Houston Methodist Hospital  Johns Hopkins Hospital  Massachusetts General Hospital  Mayo Clinic  Mount Sinai Hospital  New York-Presbyterian Columbia  New York-Presbyterian Cornell  Northwestern Memorial Hospital  Ronald Reagan UCLA Medical Center  Stanford Health Care–Stanford Hospital  The Cleveland Clinic  Tisch Hospital, NYU Langone Health  University of California, San Francisco Medical Center  University of Colorado Hospital  University of Michigan Hospitals and Health Centers  UPMC Presbyterian Shadyside, Pittsburgh  Barnes-Jewish Hospital/Washington University, St Louis  Brigham and Women’s Hospital  Cedars-Sinai Medical Center  Duke University Hospital  Hospitals of the University of Pennsylvania–Penn Presbyterian  Houston Methodist Hospital  Johns Hopkins Hospital  Massachusetts General Hospital  Mayo Clinic  Mount Sinai Hospital  New York-Presbyterian Columbia  New York-Presbyterian Cornell  Northwestern Memorial Hospital  Ronald Reagan UCLA Medical Center  Stanford Health Care–Stanford Hospital  The Cleveland Clinic  Tisch Hospital, NYU Langone Health  University of California, San Francisco Medical Center  University of Colorado Hospital  University of Michigan Hospitals and Health Centers  UPMC Presbyterian Shadyside, Pittsburgh  View Large Results Results were collected from all 21 institutions (100% response rate). Although some responses required additional communication to further clarify answers, eventually all questions were adequately addressed by all institutions. The instrument manufacturers used by the different institutions for routine CHEM and IM varied Table 3. For CHEM, only one institution used instruments from more than one vendor, whereas all other institutions used instruments from one vendor. The overall breakdown included Roche (Basel, Switzerland) (n = 10, 48%), Beckman Coulter (Brea, CA) (n = 6, 29%), Siemens (Berlin, Germany) (n = 4, 19%), and Abbott (Lake Bluff, IL) and Ortho (Raritan, NJ) (n = 1, 5% each) Table 3). In contrast, for IM, 10 (48%) of 21 institutions used instruments from multiple manufacturers (range, 1-5), including Roche (n = 13, 61.9%), Beckman Coulter (n = 10, 47.6%), Siemens (n = 9, 48.9%), Abbott (n = 7, 33.3%), Ortho (n = 1, 4.7%), and Bio-Rad (Hercules, CA) (n = 1, 4.7%). Table 3 Instruments Used for Automated Chemistry and Immunochemistry Hospital  Chemistry  Immunochemistry  A  Roche  Roche  B  Siemens  Siemens  C  Roche  Roche, Abbott  D  Siemens  Siemens  E  Roche  Roche  F  Roche, Beckman Coulter  Abbott, Roche, Beckman Coulter  G  Ortho  Ortho, Roche, Abbott  H  Beckman Coulter  Beckman Coulter, Roche  I  Beckman Coulter  Beckman Coulter, Siemens  J  Beckman Coulter  Roche, Abbott, Beckman Coulter, Siemens  K  Abbott  Abbott  L  Roche  Roche  M  Roche  Roche  N  Roche  Roche, Beckman Coulter, Siemens  O  Roche  Roche, Siemens, Beckman Coulter  P  Beckman Coulter  Beckman Coulter  Q  Roche  Roche, Abbott, Beckman Coulter  R  Siemens  Siemens  S  Siemens  Siemens  T  Roche  Roche, Siemens, Bio-Rad, Abbott, Beckman Coulter  U  Beckman Coulter  Beckman Coulter  Hospital  Chemistry  Immunochemistry  A  Roche  Roche  B  Siemens  Siemens  C  Roche  Roche, Abbott  D  Siemens  Siemens  E  Roche  Roche  F  Roche, Beckman Coulter  Abbott, Roche, Beckman Coulter  G  Ortho  Ortho, Roche, Abbott  H  Beckman Coulter  Beckman Coulter, Roche  I  Beckman Coulter  Beckman Coulter, Siemens  J  Beckman Coulter  Roche, Abbott, Beckman Coulter, Siemens  K  Abbott  Abbott  L  Roche  Roche  M  Roche  Roche  N  Roche  Roche, Beckman Coulter, Siemens  O  Roche  Roche, Siemens, Beckman Coulter  P  Beckman Coulter  Beckman Coulter  Q  Roche  Roche, Abbott, Beckman Coulter  R  Siemens  Siemens  S  Siemens  Siemens  T  Roche  Roche, Siemens, Bio-Rad, Abbott, Beckman Coulter  U  Beckman Coulter  Beckman Coulter  View Large Table 3 Instruments Used for Automated Chemistry and Immunochemistry Hospital  Chemistry  Immunochemistry  A  Roche  Roche  B  Siemens  Siemens  C  Roche  Roche, Abbott  D  Siemens  Siemens  E  Roche  Roche  F  Roche, Beckman Coulter  Abbott, Roche, Beckman Coulter  G  Ortho  Ortho, Roche, Abbott  H  Beckman Coulter  Beckman Coulter, Roche  I  Beckman Coulter  Beckman Coulter, Siemens  J  Beckman Coulter  Roche, Abbott, Beckman Coulter, Siemens  K  Abbott  Abbott  L  Roche  Roche  M  Roche  Roche  N  Roche  Roche, Beckman Coulter, Siemens  O  Roche  Roche, Siemens, Beckman Coulter  P  Beckman Coulter  Beckman Coulter  Q  Roche  Roche, Abbott, Beckman Coulter  R  Siemens  Siemens  S  Siemens  Siemens  T  Roche  Roche, Siemens, Bio-Rad, Abbott, Beckman Coulter  U  Beckman Coulter  Beckman Coulter  Hospital  Chemistry  Immunochemistry  A  Roche  Roche  B  Siemens  Siemens  C  Roche  Roche, Abbott  D  Siemens  Siemens  E  Roche  Roche  F  Roche, Beckman Coulter  Abbott, Roche, Beckman Coulter  G  Ortho  Ortho, Roche, Abbott  H  Beckman Coulter  Beckman Coulter, Roche  I  Beckman Coulter  Beckman Coulter, Siemens  J  Beckman Coulter  Roche, Abbott, Beckman Coulter, Siemens  K  Abbott  Abbott  L  Roche  Roche  M  Roche  Roche  N  Roche  Roche, Beckman Coulter, Siemens  O  Roche  Roche, Siemens, Beckman Coulter  P  Beckman Coulter  Beckman Coulter  Q  Roche  Roche, Abbott, Beckman Coulter  R  Siemens  Siemens  S  Siemens  Siemens  T  Roche  Roche, Siemens, Bio-Rad, Abbott, Beckman Coulter  U  Beckman Coulter  Beckman Coulter  View Large There was wide variation in the frequency of running CHEM QC Table 4, ranging from daily (n = 3, 14%) to every 2 hours (n = 2, 10%); intermediate intervals included every 4 hours (n = 3, 14%), 6 hours (n = 1, 5%), 8 hours (n = 6, 29%), and 12 hours (n = 6, 29%). Two (10%) institutions used a different QC frequency for electrolytes vs other CHEM tests. The total number of QC “events” per analyte (defined as the total number of times QC was run on an analyzer regardless of the number of levels) in a 24-hour period varied from a minimum of one to a maximum of 12. Three (14%) institutions used only one level per shift or one alternating level on the nonday shift hours (eg, a high control in the morning, a low control in the afternoon, and medium control on the late-night shifts). Table 4 Quality Control Frequency         QC Events/d  Hospital  CHEM  IM  Stat IM  CHEM  IM  Tn  hCG  A  2-3 Lv qd (electrolytes q8h)  2-3 Lv qd  2-3 Lv qd, negative QC q8h for Tn and hCG  3  1  3  3  B  2-3 Lv q8h  2-3 Lv q8h  2-3 Lv q8h  3  3  3  3  C  1 Lv (alternating) q2h  2-3 Lv q8h  cTn 4 Lv qd, 2 Lv alternating q2h, hCG 2 Lv q8h  12  3  12  3  D  2 Lv q12h  2 Lv q12h  2 Lv q8h  2  2  3  3  E  2 Lv q8h  2 Lv qd  2 Lv qd  3  1  1  1  F  2 Lv q8h  2 Lv qd  2 Lv qd  3  1  1  1  G  High control q12h, low control qd  2-3 Lv qd (medium Lv q12 for Lv 3 tests) or high Lv q12h/low Lv qd  Tn/CKMB/NT-proBNP: 2 Lv q12h; hCG: high q12h, low qd  2  2  2  2  H  2-3 Lv q12h  2-3 q12h  2-3 q12h  2  2  2  2  I  2-3 Lv q8h or qd  2-3 Lv qd  2-3 Lv qd  3  1  1  1  J  2 Lv q8h (q4h for electrolytes)  2-3 qd, 1 Lv (alternating) second/third shift  2-3 Lv day shift, 1 Lv (alternating) second/third shifts  6  3  3  3  K  2 Lv qd  2-3 Lv qd  Tn/hCG: 3 Lv qd  1  1  1  1  L  2 Lv q6h, 2 Lv startup/shutdown  2-3 Lv q12h  5 Lv q8h; hCG 2-3 Lv q12h  6  2  3  2  M  2 Lv at startup, 1 Lv (q4h alternating)  2-3 Lv at startup, 1 Lv at shutdown  Tn/hCG 2/3 Lv at startup, 1 Lv (alternating) q4h  7  2  7  7  N  3 Lv startup/shutdown, then 2 levels QC q2h  3 Lv startup/shutdown, plus 2 levels q8h  3 Lv q8h  12  4  3  3  O  3 Lv qd  2-3 Lv qd  Core laboratory: Tn/hCG 3 Lvq12h/q24 hours; ED laboratory: Tn/hCG 2 Lv qd  1  1  2  1  P  2-3 Lv q12h  2 Lv q8h  2 Lv q8h  2  3  3  3  Q  3 Lv day shift, 2 Lv other shifts  3 Lv day shift, 2 Lv other shifts  3 Lv qd, 2 Lv q12h  3  3  2  2  R  2 Lv q12h  2 Lv q12h  3 Lv q12h  2  2  2  2  S  2-3 q12h  q12h, certain tests once at start up  q12h, some qd  2  2  2  2  T  2 Lv qd  2 Lv qd  2 Lv qd  1  1  1  1  U  3 Lv q8h  3 Lv qd  3 Lv qd  3  1  1  1          QC Events/d  Hospital  CHEM  IM  Stat IM  CHEM  IM  Tn  hCG  A  2-3 Lv qd (electrolytes q8h)  2-3 Lv qd  2-3 Lv qd, negative QC q8h for Tn and hCG  3  1  3  3  B  2-3 Lv q8h  2-3 Lv q8h  2-3 Lv q8h  3  3  3  3  C  1 Lv (alternating) q2h  2-3 Lv q8h  cTn 4 Lv qd, 2 Lv alternating q2h, hCG 2 Lv q8h  12  3  12  3  D  2 Lv q12h  2 Lv q12h  2 Lv q8h  2  2  3  3  E  2 Lv q8h  2 Lv qd  2 Lv qd  3  1  1  1  F  2 Lv q8h  2 Lv qd  2 Lv qd  3  1  1  1  G  High control q12h, low control qd  2-3 Lv qd (medium Lv q12 for Lv 3 tests) or high Lv q12h/low Lv qd  Tn/CKMB/NT-proBNP: 2 Lv q12h; hCG: high q12h, low qd  2  2  2  2  H  2-3 Lv q12h  2-3 q12h  2-3 q12h  2  2  2  2  I  2-3 Lv q8h or qd  2-3 Lv qd  2-3 Lv qd  3  1  1  1  J  2 Lv q8h (q4h for electrolytes)  2-3 qd, 1 Lv (alternating) second/third shift  2-3 Lv day shift, 1 Lv (alternating) second/third shifts  6  3  3  3  K  2 Lv qd  2-3 Lv qd  Tn/hCG: 3 Lv qd  1  1  1  1  L  2 Lv q6h, 2 Lv startup/shutdown  2-3 Lv q12h  5 Lv q8h; hCG 2-3 Lv q12h  6  2  3  2  M  2 Lv at startup, 1 Lv (q4h alternating)  2-3 Lv at startup, 1 Lv at shutdown  Tn/hCG 2/3 Lv at startup, 1 Lv (alternating) q4h  7  2  7  7  N  3 Lv startup/shutdown, then 2 levels QC q2h  3 Lv startup/shutdown, plus 2 levels q8h  3 Lv q8h  12  4  3  3  O  3 Lv qd  2-3 Lv qd  Core laboratory: Tn/hCG 3 Lvq12h/q24 hours; ED laboratory: Tn/hCG 2 Lv qd  1  1  2  1  P  2-3 Lv q12h  2 Lv q8h  2 Lv q8h  2  3  3  3  Q  3 Lv day shift, 2 Lv other shifts  3 Lv day shift, 2 Lv other shifts  3 Lv qd, 2 Lv q12h  3  3  2  2  R  2 Lv q12h  2 Lv q12h  3 Lv q12h  2  2  2  2  S  2-3 q12h  q12h, certain tests once at start up  q12h, some qd  2  2  2  2  T  2 Lv qd  2 Lv qd  2 Lv qd  1  1  1  1  U  3 Lv q8h  3 Lv qd  3 Lv qd  3  1  1  1  CHEM, chemistry; CKMB, creatine kinase MB isoenzyme; cTn, cardiac troponin; ED, emergency department; hCG, human chorionic gonadotropin; IM, immunochemistry; Lv, level; NT-proBNP, NT-pro B-type natriuretic peptide; QC, quality control; QC Event, one or more QC levels run consecutively at a specified time on an analyzer; qd, per day; q(x)h, every x hours; Tn, troponin. View Large Table 4 Quality Control Frequency         QC Events/d  Hospital  CHEM  IM  Stat IM  CHEM  IM  Tn  hCG  A  2-3 Lv qd (electrolytes q8h)  2-3 Lv qd  2-3 Lv qd, negative QC q8h for Tn and hCG  3  1  3  3  B  2-3 Lv q8h  2-3 Lv q8h  2-3 Lv q8h  3  3  3  3  C  1 Lv (alternating) q2h  2-3 Lv q8h  cTn 4 Lv qd, 2 Lv alternating q2h, hCG 2 Lv q8h  12  3  12  3  D  2 Lv q12h  2 Lv q12h  2 Lv q8h  2  2  3  3  E  2 Lv q8h  2 Lv qd  2 Lv qd  3  1  1  1  F  2 Lv q8h  2 Lv qd  2 Lv qd  3  1  1  1  G  High control q12h, low control qd  2-3 Lv qd (medium Lv q12 for Lv 3 tests) or high Lv q12h/low Lv qd  Tn/CKMB/NT-proBNP: 2 Lv q12h; hCG: high q12h, low qd  2  2  2  2  H  2-3 Lv q12h  2-3 q12h  2-3 q12h  2  2  2  2  I  2-3 Lv q8h or qd  2-3 Lv qd  2-3 Lv qd  3  1  1  1  J  2 Lv q8h (q4h for electrolytes)  2-3 qd, 1 Lv (alternating) second/third shift  2-3 Lv day shift, 1 Lv (alternating) second/third shifts  6  3  3  3  K  2 Lv qd  2-3 Lv qd  Tn/hCG: 3 Lv qd  1  1  1  1  L  2 Lv q6h, 2 Lv startup/shutdown  2-3 Lv q12h  5 Lv q8h; hCG 2-3 Lv q12h  6  2  3  2  M  2 Lv at startup, 1 Lv (q4h alternating)  2-3 Lv at startup, 1 Lv at shutdown  Tn/hCG 2/3 Lv at startup, 1 Lv (alternating) q4h  7  2  7  7  N  3 Lv startup/shutdown, then 2 levels QC q2h  3 Lv startup/shutdown, plus 2 levels q8h  3 Lv q8h  12  4  3  3  O  3 Lv qd  2-3 Lv qd  Core laboratory: Tn/hCG 3 Lvq12h/q24 hours; ED laboratory: Tn/hCG 2 Lv qd  1  1  2  1  P  2-3 Lv q12h  2 Lv q8h  2 Lv q8h  2  3  3  3  Q  3 Lv day shift, 2 Lv other shifts  3 Lv day shift, 2 Lv other shifts  3 Lv qd, 2 Lv q12h  3  3  2  2  R  2 Lv q12h  2 Lv q12h  3 Lv q12h  2  2  2  2  S  2-3 q12h  q12h, certain tests once at start up  q12h, some qd  2  2  2  2  T  2 Lv qd  2 Lv qd  2 Lv qd  1  1  1  1  U  3 Lv q8h  3 Lv qd  3 Lv qd  3  1  1  1          QC Events/d  Hospital  CHEM  IM  Stat IM  CHEM  IM  Tn  hCG  A  2-3 Lv qd (electrolytes q8h)  2-3 Lv qd  2-3 Lv qd, negative QC q8h for Tn and hCG  3  1  3  3  B  2-3 Lv q8h  2-3 Lv q8h  2-3 Lv q8h  3  3  3  3  C  1 Lv (alternating) q2h  2-3 Lv q8h  cTn 4 Lv qd, 2 Lv alternating q2h, hCG 2 Lv q8h  12  3  12  3  D  2 Lv q12h  2 Lv q12h  2 Lv q8h  2  2  3  3  E  2 Lv q8h  2 Lv qd  2 Lv qd  3  1  1  1  F  2 Lv q8h  2 Lv qd  2 Lv qd  3  1  1  1  G  High control q12h, low control qd  2-3 Lv qd (medium Lv q12 for Lv 3 tests) or high Lv q12h/low Lv qd  Tn/CKMB/NT-proBNP: 2 Lv q12h; hCG: high q12h, low qd  2  2  2  2  H  2-3 Lv q12h  2-3 q12h  2-3 q12h  2  2  2  2  I  2-3 Lv q8h or qd  2-3 Lv qd  2-3 Lv qd  3  1  1  1  J  2 Lv q8h (q4h for electrolytes)  2-3 qd, 1 Lv (alternating) second/third shift  2-3 Lv day shift, 1 Lv (alternating) second/third shifts  6  3  3  3  K  2 Lv qd  2-3 Lv qd  Tn/hCG: 3 Lv qd  1  1  1  1  L  2 Lv q6h, 2 Lv startup/shutdown  2-3 Lv q12h  5 Lv q8h; hCG 2-3 Lv q12h  6  2  3  2  M  2 Lv at startup, 1 Lv (q4h alternating)  2-3 Lv at startup, 1 Lv at shutdown  Tn/hCG 2/3 Lv at startup, 1 Lv (alternating) q4h  7  2  7  7  N  3 Lv startup/shutdown, then 2 levels QC q2h  3 Lv startup/shutdown, plus 2 levels q8h  3 Lv q8h  12  4  3  3  O  3 Lv qd  2-3 Lv qd  Core laboratory: Tn/hCG 3 Lvq12h/q24 hours; ED laboratory: Tn/hCG 2 Lv qd  1  1  2  1  P  2-3 Lv q12h  2 Lv q8h  2 Lv q8h  2  3  3  3  Q  3 Lv day shift, 2 Lv other shifts  3 Lv day shift, 2 Lv other shifts  3 Lv qd, 2 Lv q12h  3  3  2  2  R  2 Lv q12h  2 Lv q12h  3 Lv q12h  2  2  2  2  S  2-3 q12h  q12h, certain tests once at start up  q12h, some qd  2  2  2  2  T  2 Lv qd  2 Lv qd  2 Lv qd  1  1  1  1  U  3 Lv q8h  3 Lv qd  3 Lv qd  3  1  1  1  CHEM, chemistry; CKMB, creatine kinase MB isoenzyme; cTn, cardiac troponin; ED, emergency department; hCG, human chorionic gonadotropin; IM, immunochemistry; Lv, level; NT-proBNP, NT-pro B-type natriuretic peptide; QC, quality control; QC Event, one or more QC levels run consecutively at a specified time on an analyzer; qd, per day; q(x)h, every x hours; Tn, troponin. View Large There was also variability in the frequency and number of QC for IM and STAT IM (eg, troponin and β–human chorionic gonadotropin [hCG]). The total number of non-STAT IM QC levels used per day varied between institutions from a minimum of two levels to a maximum of three different levels per day (Table 4). QC events per 24 hours were less varied than for CHEMs, ranging from one to seven. For STAT IMs, 19 laboratories used two or three levels, with one laboratory using four levels for troponin (which included a QC near the 99% cutoff), another laboratory using five levels for troponin, and four laboratories using alternating levels at different time points. QC events for STAT IM per 24 hours varied from one to 12 per day. As with CHEM, some hospitals used a single control level on select shifts. For CHEM, many hospitals used QC materials from multiple sources (n = 8, 38%) Table 5. However, all respondents predominantly relied on third-party QC materials (n = 21, 100%). Of the third-party reagents, Bio-Rad was the most common (n = 19, 90%), with MAS ChemTrak (Thermo Fisher Scientific, Waltham, MA) less commonly used (n = 2, 10%). For IM, 12 (57%) laboratories used more than one source for control materials. Again, Bio-Rad QC materials were the most commonly used (n = 19, 90%), and two (10%) laboratories favored the use of manufacturer QC materials over third party. Table 5 Quality Control Material Vendors Hospital  Chemistry  Immunochemistry  A  MAS Chemtrak  Bio-Rad  B  Bio-Rad  Bio-Rad  C  Bio-Rad  Bio-Rad  D  Bio-Rad  Bio-Rad  E  Bio-Rad, manufacturer when necessary  Bio-Rad, manufacturer QC when necessary  F  Bio-Rad  Bio-Rad  G  Bio-Rad  Bio-Rad  H  Bio-Rad  Bio-Rad  I  Predominantly Bio- Rad QC and some manufacturer supplied (if available)  Predominantly Bio-Rad QC and some manufacturer supplied (if available)  J  Bio-Rad, MAS Chemtrak  Automatic laboratory: manufacturer supplied when possible, third party when necessary (Bio-Rad). Special chemistry laboratory: Bio-Rad when possible, manufacturer when necessary; supplemented with patient pools  K  Bio-Rad  Bio-Rad, manufacturer QC when necessary  L  Bio-Rad  Bio-Rad, Roche, Randox  M  MAS ChemTrack  Manufacturer, third party when needed (Bio-Rad)  N  Mostly Bio-Rad, also Stanbio, Thermo TDM, UTAK, Elsohly  Mostly Bio-Rad, manufacturer control where necessary  O  Bio-Rad, very few manufacturers or laboratory made  Bio-Rad, very few manufacturers  P  Bio-Rad  Bio-Rad  Q  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  R  Bio-Rad  Bio-Rad  S  Bio-Rad  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  T  Bio-Rad, some Roche  Bio-Rad, some Roche, Abbott  U  Bio-Rad, very few manufacturers  Bio-Rad, very few manufacturers  Hospital  Chemistry  Immunochemistry  A  MAS Chemtrak  Bio-Rad  B  Bio-Rad  Bio-Rad  C  Bio-Rad  Bio-Rad  D  Bio-Rad  Bio-Rad  E  Bio-Rad, manufacturer when necessary  Bio-Rad, manufacturer QC when necessary  F  Bio-Rad  Bio-Rad  G  Bio-Rad  Bio-Rad  H  Bio-Rad  Bio-Rad  I  Predominantly Bio- Rad QC and some manufacturer supplied (if available)  Predominantly Bio-Rad QC and some manufacturer supplied (if available)  J  Bio-Rad, MAS Chemtrak  Automatic laboratory: manufacturer supplied when possible, third party when necessary (Bio-Rad). Special chemistry laboratory: Bio-Rad when possible, manufacturer when necessary; supplemented with patient pools  K  Bio-Rad  Bio-Rad, manufacturer QC when necessary  L  Bio-Rad  Bio-Rad, Roche, Randox  M  MAS ChemTrack  Manufacturer, third party when needed (Bio-Rad)  N  Mostly Bio-Rad, also Stanbio, Thermo TDM, UTAK, Elsohly  Mostly Bio-Rad, manufacturer control where necessary  O  Bio-Rad, very few manufacturers or laboratory made  Bio-Rad, very few manufacturers  P  Bio-Rad  Bio-Rad  Q  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  R  Bio-Rad  Bio-Rad  S  Bio-Rad  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  T  Bio-Rad, some Roche  Bio-Rad, some Roche, Abbott  U  Bio-Rad, very few manufacturers  Bio-Rad, very few manufacturers  QC, quality control. View Large Table 5 Quality Control Material Vendors Hospital  Chemistry  Immunochemistry  A  MAS Chemtrak  Bio-Rad  B  Bio-Rad  Bio-Rad  C  Bio-Rad  Bio-Rad  D  Bio-Rad  Bio-Rad  E  Bio-Rad, manufacturer when necessary  Bio-Rad, manufacturer QC when necessary  F  Bio-Rad  Bio-Rad  G  Bio-Rad  Bio-Rad  H  Bio-Rad  Bio-Rad  I  Predominantly Bio- Rad QC and some manufacturer supplied (if available)  Predominantly Bio-Rad QC and some manufacturer supplied (if available)  J  Bio-Rad, MAS Chemtrak  Automatic laboratory: manufacturer supplied when possible, third party when necessary (Bio-Rad). Special chemistry laboratory: Bio-Rad when possible, manufacturer when necessary; supplemented with patient pools  K  Bio-Rad  Bio-Rad, manufacturer QC when necessary  L  Bio-Rad  Bio-Rad, Roche, Randox  M  MAS ChemTrack  Manufacturer, third party when needed (Bio-Rad)  N  Mostly Bio-Rad, also Stanbio, Thermo TDM, UTAK, Elsohly  Mostly Bio-Rad, manufacturer control where necessary  O  Bio-Rad, very few manufacturers or laboratory made  Bio-Rad, very few manufacturers  P  Bio-Rad  Bio-Rad  Q  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  R  Bio-Rad  Bio-Rad  S  Bio-Rad  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  T  Bio-Rad, some Roche  Bio-Rad, some Roche, Abbott  U  Bio-Rad, very few manufacturers  Bio-Rad, very few manufacturers  Hospital  Chemistry  Immunochemistry  A  MAS Chemtrak  Bio-Rad  B  Bio-Rad  Bio-Rad  C  Bio-Rad  Bio-Rad  D  Bio-Rad  Bio-Rad  E  Bio-Rad, manufacturer when necessary  Bio-Rad, manufacturer QC when necessary  F  Bio-Rad  Bio-Rad  G  Bio-Rad  Bio-Rad  H  Bio-Rad  Bio-Rad  I  Predominantly Bio- Rad QC and some manufacturer supplied (if available)  Predominantly Bio-Rad QC and some manufacturer supplied (if available)  J  Bio-Rad, MAS Chemtrak  Automatic laboratory: manufacturer supplied when possible, third party when necessary (Bio-Rad). Special chemistry laboratory: Bio-Rad when possible, manufacturer when necessary; supplemented with patient pools  K  Bio-Rad  Bio-Rad, manufacturer QC when necessary  L  Bio-Rad  Bio-Rad, Roche, Randox  M  MAS ChemTrack  Manufacturer, third party when needed (Bio-Rad)  N  Mostly Bio-Rad, also Stanbio, Thermo TDM, UTAK, Elsohly  Mostly Bio-Rad, manufacturer control where necessary  O  Bio-Rad, very few manufacturers or laboratory made  Bio-Rad, very few manufacturers  P  Bio-Rad  Bio-Rad  Q  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  R  Bio-Rad  Bio-Rad  S  Bio-Rad  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  T  Bio-Rad, some Roche  Bio-Rad, some Roche, Abbott  U  Bio-Rad, very few manufacturers  Bio-Rad, very few manufacturers  QC, quality control. View Large Most hospitals used a QC rule of 2 SD (n = 16, 76%), two (10%) used variable rules (based on the test) between 2 and 3 SD, and one (5%) used a cutoff of 3 SD Table 6. Two (10%) hospitals used derivations of the Westgard rules depending on the assay. For IM, 17 (81%) hospitals chose a 2-SD rule, one (5%) chose 3 SD, and three (14%) chose some permutation of the Westgard rules. Table 6 Quality Control Rules Hospital  Chemistry  Immunochemistry  A  ±2 SD  ±2 SD  B  ±2 SD  ±2 SD  C  ±2 SD  ±2 SD  D  ±2 SD, some ±2.5 SD and ±3 SD  ±2 SD  E  ±2 SD  ±2 SD  F  ±2 SD  ±2 SD  G  ±2 SD  ±2 SD  H  ±2 SD  ±2 SD  I  Variable (1-3S, 1-3.5S, 1-4S, 1-5S, 2-2S, 2 of 3-2S, R-4S, 3-1S, 4-1S, or N-x, 1-2S, 1-2.5S)  Variable (1-3S, 1-3.5S, 1-4S, 1-5S, 2-2S, 2 of 3-2S, R-4S, 3-1S, 4-1S, or N-x, 1-2S, 1-2.5S)  J  ±2 SD  ±2 SD  K  ±2 SD  ±2 SD  L  ±2 SD  ±2 SD  M  ±2 SD  ±2 SD  N  ±2 SD  ±2 SD  O  ±2 SD  ±2 SD  P  ±2 SD  ±2 SD  Q  3 SD  3 SD  R  ±2 SD  ±2 SD  S  2-2S, 1-3S, R-4S, 10x  2-2S, 1-3S, R-4S, 10x  T  2 or 3 SD  Some 2 or 3 SD; others Westgard rules  U  ±2 SD  ±2 SD  Hospital  Chemistry  Immunochemistry  A  ±2 SD  ±2 SD  B  ±2 SD  ±2 SD  C  ±2 SD  ±2 SD  D  ±2 SD, some ±2.5 SD and ±3 SD  ±2 SD  E  ±2 SD  ±2 SD  F  ±2 SD  ±2 SD  G  ±2 SD  ±2 SD  H  ±2 SD  ±2 SD  I  Variable (1-3S, 1-3.5S, 1-4S, 1-5S, 2-2S, 2 of 3-2S, R-4S, 3-1S, 4-1S, or N-x, 1-2S, 1-2.5S)  Variable (1-3S, 1-3.5S, 1-4S, 1-5S, 2-2S, 2 of 3-2S, R-4S, 3-1S, 4-1S, or N-x, 1-2S, 1-2.5S)  J  ±2 SD  ±2 SD  K  ±2 SD  ±2 SD  L  ±2 SD  ±2 SD  M  ±2 SD  ±2 SD  N  ±2 SD  ±2 SD  O  ±2 SD  ±2 SD  P  ±2 SD  ±2 SD  Q  3 SD  3 SD  R  ±2 SD  ±2 SD  S  2-2S, 1-3S, R-4S, 10x  2-2S, 1-3S, R-4S, 10x  T  2 or 3 SD  Some 2 or 3 SD; others Westgard rules  U  ±2 SD  ±2 SD  View Large Table 6 Quality Control Rules Hospital  Chemistry  Immunochemistry  A  ±2 SD  ±2 SD  B  ±2 SD  ±2 SD  C  ±2 SD  ±2 SD  D  ±2 SD, some ±2.5 SD and ±3 SD  ±2 SD  E  ±2 SD  ±2 SD  F  ±2 SD  ±2 SD  G  ±2 SD  ±2 SD  H  ±2 SD  ±2 SD  I  Variable (1-3S, 1-3.5S, 1-4S, 1-5S, 2-2S, 2 of 3-2S, R-4S, 3-1S, 4-1S, or N-x, 1-2S, 1-2.5S)  Variable (1-3S, 1-3.5S, 1-4S, 1-5S, 2-2S, 2 of 3-2S, R-4S, 3-1S, 4-1S, or N-x, 1-2S, 1-2.5S)  J  ±2 SD  ±2 SD  K  ±2 SD  ±2 SD  L  ±2 SD  ±2 SD  M  ±2 SD  ±2 SD  N  ±2 SD  ±2 SD  O  ±2 SD  ±2 SD  P  ±2 SD  ±2 SD  Q  3 SD  3 SD  R  ±2 SD  ±2 SD  S  2-2S, 1-3S, R-4S, 10x  2-2S, 1-3S, R-4S, 10x  T  2 or 3 SD  Some 2 or 3 SD; others Westgard rules  U  ±2 SD  ±2 SD  Hospital  Chemistry  Immunochemistry  A  ±2 SD  ±2 SD  B  ±2 SD  ±2 SD  C  ±2 SD  ±2 SD  D  ±2 SD, some ±2.5 SD and ±3 SD  ±2 SD  E  ±2 SD  ±2 SD  F  ±2 SD  ±2 SD  G  ±2 SD  ±2 SD  H  ±2 SD  ±2 SD  I  Variable (1-3S, 1-3.5S, 1-4S, 1-5S, 2-2S, 2 of 3-2S, R-4S, 3-1S, 4-1S, or N-x, 1-2S, 1-2.5S)  Variable (1-3S, 1-3.5S, 1-4S, 1-5S, 2-2S, 2 of 3-2S, R-4S, 3-1S, 4-1S, or N-x, 1-2S, 1-2.5S)  J  ±2 SD  ±2 SD  K  ±2 SD  ±2 SD  L  ±2 SD  ±2 SD  M  ±2 SD  ±2 SD  N  ±2 SD  ±2 SD  O  ±2 SD  ±2 SD  P  ±2 SD  ±2 SD  Q  3 SD  3 SD  R  ±2 SD  ±2 SD  S  2-2S, 1-3S, R-4S, 10x  2-2S, 1-3S, R-4S, 10x  T  2 or 3 SD  Some 2 or 3 SD; others Westgard rules  U  ±2 SD  ±2 SD  View Large When a QC was out of control, all but one hospital elected to repeat the control and accept results if the repeat came back into control (n = 20, 95%), although some had some minor variations (such as rejecting a run outright if a control was out by >4 SD; Table 7). One (5%) institution rejected runs if two controls were out 2 SD or if one was out 3 SD. Another institution (5%) repeated QC if one of the two levels was out 2 SD from target value but accepted the run if only one of three QC levels was out. This institution also rejected outright if the following rules were broken: 1- 2.5S, 2-2S, 2/3-2S, or R-4S. Only laboratory E made no provisions for repeating control(s); if one level exceeded 3 SD or two controls exceeded 2 SD, the run was rejected as being out of control. Table 7 Quality Control Rules Hospital  QC Rules  A  If control is out of range, it is repeated. If repeat is in range, the results are accepted. If still out of range, assay is recalibrated and/or additional troubleshooting occurs (eg, new reagent pack, new QC). Patient results may be repeated if it is determined that QC was out when patient results were reported.  B  As “A.”  C  As “A.”  D  As “A.”  E  Reject results if two QC values are out 2 SD or more or if one is out 3 SD. If QC fails, corrective action is taken. If one QC is out 2 SD, evaluate other QC in same run and in previous runs (warning only).  F  As “A.”  G  As “A.”  H  As “A” if one QC out <4 SD. If two QCs are out 2 to 4 SD, both must be repeated and corrective action is taken if one or more are still out. If QC out >4 SD, stop analysis immediately and take corrective action before continuing testing.  I  As “A.”  J  As “A”  K  As “A,” but if the QC results are acceptable after recalibration, 10 specimens analyzed within the last 24 hours are repeated. If they are within 10% of reported value, the assay is validated. If not, further corrective action is taken.  L  As “A.”  M  As “A.”  N  As “A.”  O  For analytes with two QC levels and troponin, as “A” but also rejected if 1-2.5S, 2-2S, 2/3-2S, or R-4S rules are violated. We also have a warning with the 7-T rule. For analytes with three QC levels, if control is outside of 2 SD range (1-2S), it is a warning; repeat is not required and run is accepted. If two QC levels are outside of 2 SD range (2-2S), run is rejected, and controls are repeated using new vials/aliquots of controls. If QC is back into range, the run is accepted. If repeat is out of range, investigate/recalibrate. Reject run also if 1-2.5S, 2-2S, 2/3-2S, or R-4S rule is violated. We also have a warning with the 7-T rule.  P  As “A.”  Q  If QC is out 3 SD, then QC failure; repeat once, and if it fails again, take corrective action.  R  We use several Westgard rules (4-1S, 10x) to further investigate when a 2-SD flag occurs on QC. First step is to repeat QC; if it is in and no Westgard rules have triggered, accept and continue. If out, check QC on other analyzers and rerun at least five patients on a different analyzer. Notify supervisor if recalibration is warranted.  S  If results are out by 2-2S or 1-3S, the QC is repeated and accepted if it comes in. If it is out, the system is recalibrated and QC is rerun. Any further problems are escalated to the lead in the area for further troubleshooting.  T  As “A.”  U  As “A”; if far outside of 2 SD, then may be worked up at coordinator’s discretion.  Hospital  QC Rules  A  If control is out of range, it is repeated. If repeat is in range, the results are accepted. If still out of range, assay is recalibrated and/or additional troubleshooting occurs (eg, new reagent pack, new QC). Patient results may be repeated if it is determined that QC was out when patient results were reported.  B  As “A.”  C  As “A.”  D  As “A.”  E  Reject results if two QC values are out 2 SD or more or if one is out 3 SD. If QC fails, corrective action is taken. If one QC is out 2 SD, evaluate other QC in same run and in previous runs (warning only).  F  As “A.”  G  As “A.”  H  As “A” if one QC out <4 SD. If two QCs are out 2 to 4 SD, both must be repeated and corrective action is taken if one or more are still out. If QC out >4 SD, stop analysis immediately and take corrective action before continuing testing.  I  As “A.”  J  As “A”  K  As “A,” but if the QC results are acceptable after recalibration, 10 specimens analyzed within the last 24 hours are repeated. If they are within 10% of reported value, the assay is validated. If not, further corrective action is taken.  L  As “A.”  M  As “A.”  N  As “A.”  O  For analytes with two QC levels and troponin, as “A” but also rejected if 1-2.5S, 2-2S, 2/3-2S, or R-4S rules are violated. We also have a warning with the 7-T rule. For analytes with three QC levels, if control is outside of 2 SD range (1-2S), it is a warning; repeat is not required and run is accepted. If two QC levels are outside of 2 SD range (2-2S), run is rejected, and controls are repeated using new vials/aliquots of controls. If QC is back into range, the run is accepted. If repeat is out of range, investigate/recalibrate. Reject run also if 1-2.5S, 2-2S, 2/3-2S, or R-4S rule is violated. We also have a warning with the 7-T rule.  P  As “A.”  Q  If QC is out 3 SD, then QC failure; repeat once, and if it fails again, take corrective action.  R  We use several Westgard rules (4-1S, 10x) to further investigate when a 2-SD flag occurs on QC. First step is to repeat QC; if it is in and no Westgard rules have triggered, accept and continue. If out, check QC on other analyzers and rerun at least five patients on a different analyzer. Notify supervisor if recalibration is warranted.  S  If results are out by 2-2S or 1-3S, the QC is repeated and accepted if it comes in. If it is out, the system is recalibrated and QC is rerun. Any further problems are escalated to the lead in the area for further troubleshooting.  T  As “A.”  U  As “A”; if far outside of 2 SD, then may be worked up at coordinator’s discretion.  QC, quality control. View Large Table 7 Quality Control Rules Hospital  QC Rules  A  If control is out of range, it is repeated. If repeat is in range, the results are accepted. If still out of range, assay is recalibrated and/or additional troubleshooting occurs (eg, new reagent pack, new QC). Patient results may be repeated if it is determined that QC was out when patient results were reported.  B  As “A.”  C  As “A.”  D  As “A.”  E  Reject results if two QC values are out 2 SD or more or if one is out 3 SD. If QC fails, corrective action is taken. If one QC is out 2 SD, evaluate other QC in same run and in previous runs (warning only).  F  As “A.”  G  As “A.”  H  As “A” if one QC out <4 SD. If two QCs are out 2 to 4 SD, both must be repeated and corrective action is taken if one or more are still out. If QC out >4 SD, stop analysis immediately and take corrective action before continuing testing.  I  As “A.”  J  As “A”  K  As “A,” but if the QC results are acceptable after recalibration, 10 specimens analyzed within the last 24 hours are repeated. If they are within 10% of reported value, the assay is validated. If not, further corrective action is taken.  L  As “A.”  M  As “A.”  N  As “A.”  O  For analytes with two QC levels and troponin, as “A” but also rejected if 1-2.5S, 2-2S, 2/3-2S, or R-4S rules are violated. We also have a warning with the 7-T rule. For analytes with three QC levels, if control is outside of 2 SD range (1-2S), it is a warning; repeat is not required and run is accepted. If two QC levels are outside of 2 SD range (2-2S), run is rejected, and controls are repeated using new vials/aliquots of controls. If QC is back into range, the run is accepted. If repeat is out of range, investigate/recalibrate. Reject run also if 1-2.5S, 2-2S, 2/3-2S, or R-4S rule is violated. We also have a warning with the 7-T rule.  P  As “A.”  Q  If QC is out 3 SD, then QC failure; repeat once, and if it fails again, take corrective action.  R  We use several Westgard rules (4-1S, 10x) to further investigate when a 2-SD flag occurs on QC. First step is to repeat QC; if it is in and no Westgard rules have triggered, accept and continue. If out, check QC on other analyzers and rerun at least five patients on a different analyzer. Notify supervisor if recalibration is warranted.  S  If results are out by 2-2S or 1-3S, the QC is repeated and accepted if it comes in. If it is out, the system is recalibrated and QC is rerun. Any further problems are escalated to the lead in the area for further troubleshooting.  T  As “A.”  U  As “A”; if far outside of 2 SD, then may be worked up at coordinator’s discretion.  Hospital  QC Rules  A  If control is out of range, it is repeated. If repeat is in range, the results are accepted. If still out of range, assay is recalibrated and/or additional troubleshooting occurs (eg, new reagent pack, new QC). Patient results may be repeated if it is determined that QC was out when patient results were reported.  B  As “A.”  C  As “A.”  D  As “A.”  E  Reject results if two QC values are out 2 SD or more or if one is out 3 SD. If QC fails, corrective action is taken. If one QC is out 2 SD, evaluate other QC in same run and in previous runs (warning only).  F  As “A.”  G  As “A.”  H  As “A” if one QC out <4 SD. If two QCs are out 2 to 4 SD, both must be repeated and corrective action is taken if one or more are still out. If QC out >4 SD, stop analysis immediately and take corrective action before continuing testing.  I  As “A.”  J  As “A”  K  As “A,” but if the QC results are acceptable after recalibration, 10 specimens analyzed within the last 24 hours are repeated. If they are within 10% of reported value, the assay is validated. If not, further corrective action is taken.  L  As “A.”  M  As “A.”  N  As “A.”  O  For analytes with two QC levels and troponin, as “A” but also rejected if 1-2.5S, 2-2S, 2/3-2S, or R-4S rules are violated. We also have a warning with the 7-T rule. For analytes with three QC levels, if control is outside of 2 SD range (1-2S), it is a warning; repeat is not required and run is accepted. If two QC levels are outside of 2 SD range (2-2S), run is rejected, and controls are repeated using new vials/aliquots of controls. If QC is back into range, the run is accepted. If repeat is out of range, investigate/recalibrate. Reject run also if 1-2.5S, 2-2S, 2/3-2S, or R-4S rule is violated. We also have a warning with the 7-T rule.  P  As “A.”  Q  If QC is out 3 SD, then QC failure; repeat once, and if it fails again, take corrective action.  R  We use several Westgard rules (4-1S, 10x) to further investigate when a 2-SD flag occurs on QC. First step is to repeat QC; if it is in and no Westgard rules have triggered, accept and continue. If out, check QC on other analyzers and rerun at least five patients on a different analyzer. Notify supervisor if recalibration is warranted.  S  If results are out by 2-2S or 1-3S, the QC is repeated and accepted if it comes in. If it is out, the system is recalibrated and QC is rerun. Any further problems are escalated to the lead in the area for further troubleshooting.  T  As “A.”  U  As “A”; if far outside of 2 SD, then may be worked up at coordinator’s discretion.  QC, quality control. View Large Although most of the surveyed hospitals do not currently use moving averages (n = 19, 90%), four (19%) are hoping to implement moving averages in the near future Table 8. One (5%) institution runs moving averages in the background but does not use them for clinical metrics. Only one (5%) uses moving averages for clinical use and then only for a small number of assays. One of the institutions that did not use moving averages reported that it had previously implemented them but had not found them to be useful. Table 8 Moving Averages Hospital  Do You Use Moving Averages?  A  No  B  No  C  No  D  No  E  No  F  Not yet, but planning to  G  No  H  No  I  No  J  No [but hopefully soon]  K  No  L  Not clinically, run for some in background to collect data  M  No  N  No, investigated but not useful  O  No  P  No  Q  No  R  We do use moving averages routinely as QC point on six different assays and have the capability of turning it on for other assays if deemed necessary.  S  We are collecting data currently and will be implementing spring/summer of 2017.  T  No  U  Not at this time, but planning to implement this year  Hospital  Do You Use Moving Averages?  A  No  B  No  C  No  D  No  E  No  F  Not yet, but planning to  G  No  H  No  I  No  J  No [but hopefully soon]  K  No  L  Not clinically, run for some in background to collect data  M  No  N  No, investigated but not useful  O  No  P  No  Q  No  R  We do use moving averages routinely as QC point on six different assays and have the capability of turning it on for other assays if deemed necessary.  S  We are collecting data currently and will be implementing spring/summer of 2017.  T  No  U  Not at this time, but planning to implement this year  QC, quality control. View Large Table 8 Moving Averages Hospital  Do You Use Moving Averages?  A  No  B  No  C  No  D  No  E  No  F  Not yet, but planning to  G  No  H  No  I  No  J  No [but hopefully soon]  K  No  L  Not clinically, run for some in background to collect data  M  No  N  No, investigated but not useful  O  No  P  No  Q  No  R  We do use moving averages routinely as QC point on six different assays and have the capability of turning it on for other assays if deemed necessary.  S  We are collecting data currently and will be implementing spring/summer of 2017.  T  No  U  Not at this time, but planning to implement this year  Hospital  Do You Use Moving Averages?  A  No  B  No  C  No  D  No  E  No  F  Not yet, but planning to  G  No  H  No  I  No  J  No [but hopefully soon]  K  No  L  Not clinically, run for some in background to collect data  M  No  N  No, investigated but not useful  O  No  P  No  Q  No  R  We do use moving averages routinely as QC point on six different assays and have the capability of turning it on for other assays if deemed necessary.  S  We are collecting data currently and will be implementing spring/summer of 2017.  T  No  U  Not at this time, but planning to implement this year  QC, quality control. View Large Discussion QC is a critical aspect of laboratory management that laboratory directors take very seriously as it helps ensure we provide accurate results to guide clinical management. This is perhaps related to the high rate of response (100%) from our cohort. In this study, we were able to survey a wide range of highly regarded academic institutions comprising the entirety of the US News & World Report 2016 to 2017 honor roll list.7 This list consisted of 21 hospitals from 12 states. We selected this cohort as a group of high-performing academic institutions with diverse practice settings and distinct academic histories. As expected, a variety of different instruments were used in the laboratories surveyed. Although some vendors were seen more frequently among respondents, no single vendor or platform had complete market dominance. Since QC rules are often related to actual assay performance, it makes sense that different platforms, which will have distinct methods, will perform somewhat differently and may require different QC rules. There were no apparent trends between QC frequency and manufacturer, and variation existed between laboratories using similar instruments. Therefore, it seems unlikely that platform choice alone would affect QC practices to a significant extent. A factor that supports this is that, in most institutions that used instruments from multiple manufacturers, QC rules did not vary notably between the instruments. There was dramatic (ie, 12-fold) variation in QC frequency, ranging from once daily to every 2 hours. This was surprising because, although QC frequency may vary based on the device used, reagent stability, and test volume, the clinical risk associated with result errors should be more or less similar among the cohort, especially for routine CHEM/IM testing. Possible factors include assay method, test volume, economic constraints, and the difficulty of repeating patient samples in the event of QC failure. When QC fails, there is the possibility of needing to repeat all patient samples tested since the previous successful QC. For a high-volume test, potentially repeating all samples from the previous 24 hours would result in a very large number of samples that must be retrieved and retested. In addition, a significant delay before erroneous results are corrected has the potential to affect patient care. These are reasonable concerns, but increased frequency of QC may result in increased QC failures for purely statistical reasons, resulting in unnecessary corrective actions, delayed results, and increased costs. Other QC practices were reported by several laboratories in the survey. At the time of surveying, three laboratories (C, J, and M) performed “alternating-level” QC testing—that is, instead of testing two levels of QC material at t = X hours after daily startup, one level was tested at t = X/2 hours and an alternate level tested at t = X hours. Laboratories C and M performed such “alternating” QC testing for CHEMs, all three for STAT IM (laboratory C did for troponin but not for hCG), and only laboratory J for IMs (in the time since surveying, laboratory J has ceased performing alternating-level QC). In theory, this practice reduces the time to detect an out-of-control situation compared with performing both QC levels at double the time interval. This practice assumes that all out-of-control situations will be detected equally well by testing either QC level, an assumption that may not apply in all analytical circumstances. Several laboratories reduced the number of QC levels tested after daily startup; for example, if three levels of QC were tested to demonstrate an instrument test was “in control” at time t = 0 hours, subsequent QC testing events might only employ one or two QC levels over the subsequent 24 hours. Four laboratories (G, M, N, and Q) did this for CHEM, five for IM (G, J, M, N, and Q), and six for STAT IM (A, C [troponin only], G [hCG only], J, M, and Q). Two laboratories (A and J) performed QC testing for electrolytes (Na, K, and Cl) at a higher frequency than CHEM, likely due to the high volume of tests. QC materials were overwhelmingly third party, with only a single laboratory relying predominantly on manufacturer-supplied QC materials (for IM only). Third-party materials have the theoretical advantage of providing a more independent verification of assay function and may have the option of QC material return to the parent company for further analysis if repeated failure occurs with a lot. On the other hand, manufacturer-produced materials have the potential benefit of being specifically designed for the system and test in question such that assay failures can be traced to a single manufacturer source, as opposed to having to query two separate vendors for QC material and machine issues. Despite these theoretical considerations, there has not been a systematic study into which method provides the most reliable approach for the selection of QC materials. Most (n = 16, 76%) respondents used a QC range of 2 SD almost exclusively, and 14% (n = 3) used a combination of between 2 and 3 SD. This is an unexpected finding, as there is no standard Westgard rule for 1-2S, except as a warning. Assuming a normal distribution of values for QC materials on repeat analysis (ie, normal random variation and no systemic bias), a QC will be out of a 2-SD control range approximately 5% of the time compared with only 1% if a cutoff of 3 SD is used. Although a 2-SD QC rule has an increased chance of finding small analytical variations, one would expect lowered specificity, with numerous incidences of QC out-of-range results merely due to chance. In contrast, in laboratories where 3 SD or 2 × 2 SD are used, one would theoretically expect a potential 1% or 0.25% rate of QC out of control based on random error. Of the 21 hospitals that responded to the survey, only two explicitly used Westgard rule derivations. These rules were introduced in the 1970s by James Westgard and colleagues in an effort to apply a mathematically rigorous approach to systematizing quality control. The Westgard rules are used to evaluate QC data and are designed to capture both increased variation (random error) and bias (systemic error) while minimizing both false negatives and false positives.8 Minimization of false negatives is especially critical, since these patients may be entirely missed or discharged without proper workup (as opposed to false positives, which may be revelated by further testing). There are several iterations of the Westgard rules, but the process of flagging and repeating controls out at a 2-SD level is not one of the standard criteria. Assuming normal variation, approximately one in 20 QC results will be out of range, a high level of false positives. However, if run rejection requires two consecutive 2-SD errors, the base possibility of rejection is 0.25%, a lower percentage than that of the 3-SD cutoff. Relying on a QC to be abnormal and then repeat as abnormal requires a high level of systemic bias, and if the issue is increased variability, a repeated QC could very well be normal. Even if there is bias, the requirement for two consecutive QC measurements to be out at 2 SD may miss many low-lying biases that a 10x rule (which flags a run if 10 consecutive QC results are off by 1 SD in the same direction) would detect. In contrast, an increase in assay variability may be detected better by a 1-3S rule, which would flag a run where a single QC value is out by 3 SD. Of course, combining both of these rules would allow for detection of these two very different types of error. In this manner, multirule QC checks have the capacity to detect different errors at a higher sensitivity than a single blanket rule. Although not a large number of studies have validated the practice of repeating QC samples, a study in 2012 suggested that this process can provide performance on par with a 1-3S/2-2S/R-4S multirule, with the tradeoff of slightly increased cost of QC materials (due to increased rate of repetition).9 However, this study was limited by the use of a simple in silico model (that simulated differing levels of systemic error) and only compared with an abbreviated Westgard multirule. An additional level of variation that was not captured in this survey is the method of SD derivation. The most common method for calculating SD for QC purposes is to run a control analyte numerous times (usually at least 20 times) and measure the SD of the results. Depending on the laboratory, machine, and analyte, the number of repeats used to calculate the SD may vary. In addition, many controls have manufacturer-recommended standard deviation ranges as part of their package insert. The overwhelming majority of hospitals did not use moving averages (90%), although there was interest in implementing this in 19% of the hospitals. Only one hospital used moving averages as part of its QC practice (the other ran them in the background but did not fully use them), and one hospital had previously used moving averages but discontinued them due to perceived lack of utility. Overall, this is a surprising finding, as moving averages are theoretically useful, as well as inexpensive and simple to implement. The software to track these moving averages is included in commercially available software systems. Moving averages have the potential to detect low-level drifts in the values of measurements that would not ordinarily trigger normal QC flags.10-13 One laboratory mentioned that it used performance-driven QC methods, another proposed means of establishing QC goals.14 This involves using biological variation data to set goals for total allowable error and QC rules. For tests such as alanine aminotransferase, which have a high intraindividual and interindividual variation (24.3% and 41.6%, respectively),15 the goals for acceptable imprecision could theoretically be relaxed from ±2 SD to ±2.5 or even ±3 SD. This method may result in a reduction in the number of false rejections and maximize the number of true rejects. Overall, our findings demonstrate a heterogenous but surprisingly similar grouping of QC practices at these academic laboratories. At least 75% of the hospitals used a QC range of 2 SD, and virtually all (90%) used the policy of repeating an out-of-control QC and accepting results if the repeat value comes into control. The method of repeating out-of-control QC at 2 SD has been shown to be effective at improving the performance of QC over simple multirule methods (in this case, 1-3SD, 2 of 3-2SD/R-4) in an in silico model.9 In comparison to these QC methods developed through mathematical analysis, most of these policies appear to have evolved in a fashion “validated by experience.” The 2 SD or similar cutoff with QC repetition was seen commonly in our cohort, demonstrating that it is likely deeply ingrained in clinical pathology. Limitations of this study include the survey-based method: to ensure maximal compliance, we limited the number of items that we queried. Extremely detailed and granular information for each institution was not always available, so the data set used was not uniform, which creates the possibility of misinterpretation based on the answers. We attempted to minimize this by circulating the manuscript with the participating authors (each author was aware of which anonymized letter corresponded to his or her laboratory). Total quality management of a laboratory includes QC as a basic tenet but also a wide array of next-level practices to ensure test result quality. Future directions for this research include deeper-level inquiry into the heterogeneity of QC programs as well as investigating how these different practices affect result reporting and patient care. In conclusion, this study demonstrated both similarities and differences among QC practices at academic hospitals. There appears to be no systematic approach to defining QC rules or frequency. The Westgard rules offer a systematic and thoroughly vetted approach for QC to detect errors while minimizing false-positive rates, and additional methods involving rules for QC-level repetition have also been mathematically studied. Interestingly, Westgard rules were used by a small minority of academic center laboratories. Most laboratories prefer home-validated QC rules, which may rely on the director’s experience and expertise rather than a rigorously validated statistical approach to designing QC rules. We believe that most academic medical center chemistry laboratories that have similar volumes and patient populations may benefit from a standardized approach to QC for routine chemistry/immunochemistry testing and that can withstand rigorous statistical validation. The survey results suggest an opportunity for laboratory professional organizations to convene a consensus panel to determine a best practice approach (or approaches) to QC in the chemistry laboratory. This would help to ensure quality testing by enhancing error detection and reduce the costs associated with excessive QC and the use of QC rules that create high (false-positive) run rejection rates. References 1. Clinical Laboratory Improvement Amendments (CLIA). https://wwwn.cdc.gov/clia/Regulatory/default.aspx. Accessed January 1, 2017. 2. Kaplan AK,Pesce AJ. Clinical Chemistry: Theory, Analysis, Correlation . St Louis, MO: Mosby; 2009. 3. Burtis C, Ashwood E, Bruns D. Tietz Textbook of Clinical Chemistry and Molecular Diagnostics . Amsterdam, the Netherlands: Elsevier Saunders; 2006. 4. Mcpherson RA, Pincus MR. Henrys Clinical Diagnosis and Management by Laboratory Methods . Philadelphia, PA: Saunders; 2011. 5. Westgard JO, Barry PL, Hunt MRet al.   A multi-rule Shewhart chart for quality control in clinical chemistry. Clin Chem . 1981; 27: 493- 501. Google Scholar PubMed  6. Coskun A. Westgard multirule for calculated laboratory tests. Clin Chem Lab Med . 2006; 44: 1183- 1187. Google Scholar PubMed  7. Harder AC, Ben. 2016-17 Best hospitals honor roll and overview. 2016. http://health.usnews.com/health-care/best-hospitals/articles/best-hospitals-honor-roll-and-overview. Accessed January 1, 2017. 8. Westgard JO, Groth T, Aronsson Tet al.   Performance characteristics of rules for internal quality control: probabilities for false rejection and error detection. Clin Chem . 1977; 23: 1857- 1867. Google Scholar PubMed  9. Parvin CA, Kuchipudi L, Yundt-Pacheco JC. Should I repeat my 1:2s QC rejection? Clin Chem . 2012; 58: 925- 929. Google Scholar CrossRef Search ADS PubMed  10. Fleming JK, Katayev A. Changing the paradigm of laboratory quality control through implementation of real-time test results monitoring: for patients by patients. Clin Biochem . 2015; 48: 508- 513. Google Scholar CrossRef Search ADS PubMed  11. Wilson A, Roberts WL, Pavlov Iet al.   Patient result median monitoring for clinical laboratory quality control. Clin Chim Acta . 2011; 412: 1441- 1446. Google Scholar CrossRef Search ADS PubMed  12. Liu J, Tan CH, Badrick Tet al.   Moving sum of number of positive patient result as a quality control tool. Clin Chem Lab Med . 2017; 55: 1709- 1714. Google Scholar PubMed  13. Ng D, Polito FA, Cervinski MA. Optimization of a moving averages program using a simulated annealing algorithm: the goal is to monitor the process not the patients. Clin Chem . 2016; 62: 1361- 1371. Google Scholar CrossRef Search ADS PubMed  14. Brooks ZC. Performance-Driven Quality Control . Washington, DC: AACC Press; 2001. 15. Fraser CG. Biological Variation: From Principles to Practice. Washington, DC: AACC Press; 2001. © American Society for Clinical Pathology, 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png American Journal of Clinical Pathology Oxford University Press

Loading next page...
 
/lp/ou_press/quality-control-practices-for-chemistry-and-immunochemistry-in-a-Yb00iDe5pi
Publisher
Oxford University Press
Copyright
© American Society for Clinical Pathology, 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
0002-9173
eISSN
1943-7722
D.O.I.
10.1093/ajcp/aqy033
Publisher site
See Article on Publisher Site

Abstract

Abstract Objectives In the United States, minimum standards for quality control (QC) are specified in federal law under the Clinical Laboratory Improvement Amendment and its revisions. Beyond meeting this required standard, laboratories have flexibility to determine their overall QC program. Methods We surveyed chemistry and immunochemistry QC procedures at 21 clinical laboratories within leading academic medical centers to assess if standardized QC practices exist for chemistry and immunochemistry testing. Results We observed significant variation and unexpected similarities in practice across laboratories, including QC frequency, cutoffs, number of levels analyzed, and other features. Conclusions This variation in practice indicates an opportunity exists to establish an evidence-based approach to QC that can be generalized across institutions. Quality control, Westgard rules, Quality control rules, Chemistry, Immunochemistry In order for a clinical test to be useful, it must be both accurate (reflect the actual concentration of the analyte in question) and precise (be reproducible). Quality control (QC) testing using measurements of QC samples (with known target analyte values) is a key component of the overall quality management process. In the United States, minimum acceptable standards for QC are specified in federal law under the Clinical Laboratory Improvement Amendment and its subsequent revisions.1 These standards are enforced through the laboratory accreditation process by the College of American Pathologists, The Joint Commission, and other organizations. The basic theoretical framework involved in the design of a laboratory QC program has been described in standard textbooks.2-4 These sources often emphasize the use of power function curves (which chart the probability of rejection vs the degree of systematic error for different QC rules) to determine the number and frequency of controls, as well as cutoff values.2 Furthermore, these power curves can also be used to calculate the probability of false negatives (accepting an out-of-control QC as valid) and false positives (rejecting QC that is actually valid), as well as to estimate the number of control measurements required to identify errors of varying magnitude.2 Some laboratories use a variation of the Westgard Multirule Chart.5,6 This chart uses a series of control rules that are employed to interpret QC data to determine if a result is either in or out of control. The rules were designed to be sensitive to both random and systematic errors.3 Typically, the rules are applied to a Levey-Jennings QC plot with control limits drawn to indicate varying degrees of deviation from the expected QC mean.6 Given that this statistically robust approach to QC has existed for several decades, one would expect that a consensus evidence-based best practice guideline for quality control in clinical chemistry (CHEM) and immunochemistry (IM) would have been developed during this time. However, to our knowledge, no national pathology/laboratory professional organization has issued detailed guidance statements regarding QC programs. Logically, one may assume that similar laboratories performing the same tests on comparable types of patients would have similar approaches to QC with minor variations due to differing volumes, instruments, and local practices. We surveyed 21 leading academic medical centers regarding their QC approaches to assess similarities and differences in QC practice across different organizations. Materials and Methods A six-question survey about QC practices Table 1 was distributed to laboratory directors at 21 large academic medical centers Table 2. These centers were chosen for their size and national reputation based on the US News & World Report 2016 to 2017 hospital top 20 honor roll.7 Because of their separate locations and laboratory management schemas, NewYork-Presbyterian Cornell and Columbia were considered separate, bringing the total to 21 instead of 20. For the purpose of this analysis, the results from individual laboratories were kept anonymous such that the QC program used at individual centers would not be identified. The questions were deliberately phrased in an open-ended fashion to collect as much data possible, with additional communication via email used to elucidate any ambiguities. The responses of the laboratories were tabulated to give a semiquantitative picture of QC practices among the participating hospitals. The instruments used were stratified by vendor company, with QC frequency stratified by number of QC levels and events per day (when “per shift” was used, three shifts/day were assumed, and startup and shutdown were considered independent time points). QC materials were divided into manufacturer (ie, provided by the same vendor that produces the testing platform) and third party (which was further substratified by vendor). QC rules were generally defined as a number of standard deviations (SDs) of difference from the mean acceptable to call a value “in control” (ie, a 2-SD rule means that a QC value is considered “out of control” if it is 2 SD or more from the accepted average). We did not delineate how these SDs were derived (eg, from testing a QC sample multiple times to determine the SD vs manufacturer-designated SD ranges). QC rules were generally in the format of “if test X is out by Y SD, we take action Z.” These rules were simplified and streamlined to give short, easily interpretable responses of uniform format (these changes were supplied back to the laboratories to confirm accuracy). Finally, the utilization of moving averages was coded as yes or no, and any narrative comments (eg, utilization for nonclinical purposes, intent on implementation in the near future) were considered separately and mentioned where relevant. Table 1 Questions Included on Surveys What instrument do you use for automated chemistry and immunochemistry?  For chemistry and immunochemistry, how do you perform QC (number of levels and frequency)?  What QC material do you use (manufacturer supplied or third party)?  How do you define your QC ranges?  What are your QC rules (eg, Westgard rules, other)?  Do you use moving averages?  What instrument do you use for automated chemistry and immunochemistry?  For chemistry and immunochemistry, how do you perform QC (number of levels and frequency)?  What QC material do you use (manufacturer supplied or third party)?  How do you define your QC ranges?  What are your QC rules (eg, Westgard rules, other)?  Do you use moving averages?  QC, quality control. View Large Table 1 Questions Included on Surveys What instrument do you use for automated chemistry and immunochemistry?  For chemistry and immunochemistry, how do you perform QC (number of levels and frequency)?  What QC material do you use (manufacturer supplied or third party)?  How do you define your QC ranges?  What are your QC rules (eg, Westgard rules, other)?  Do you use moving averages?  What instrument do you use for automated chemistry and immunochemistry?  For chemistry and immunochemistry, how do you perform QC (number of levels and frequency)?  What QC material do you use (manufacturer supplied or third party)?  How do you define your QC ranges?  What are your QC rules (eg, Westgard rules, other)?  Do you use moving averages?  QC, quality control. View Large Table 2 Hospitals Responding in Full Barnes-Jewish Hospital/Washington University, St Louis  Brigham and Women’s Hospital  Cedars-Sinai Medical Center  Duke University Hospital  Hospitals of the University of Pennsylvania–Penn Presbyterian  Houston Methodist Hospital  Johns Hopkins Hospital  Massachusetts General Hospital  Mayo Clinic  Mount Sinai Hospital  New York-Presbyterian Columbia  New York-Presbyterian Cornell  Northwestern Memorial Hospital  Ronald Reagan UCLA Medical Center  Stanford Health Care–Stanford Hospital  The Cleveland Clinic  Tisch Hospital, NYU Langone Health  University of California, San Francisco Medical Center  University of Colorado Hospital  University of Michigan Hospitals and Health Centers  UPMC Presbyterian Shadyside, Pittsburgh  Barnes-Jewish Hospital/Washington University, St Louis  Brigham and Women’s Hospital  Cedars-Sinai Medical Center  Duke University Hospital  Hospitals of the University of Pennsylvania–Penn Presbyterian  Houston Methodist Hospital  Johns Hopkins Hospital  Massachusetts General Hospital  Mayo Clinic  Mount Sinai Hospital  New York-Presbyterian Columbia  New York-Presbyterian Cornell  Northwestern Memorial Hospital  Ronald Reagan UCLA Medical Center  Stanford Health Care–Stanford Hospital  The Cleveland Clinic  Tisch Hospital, NYU Langone Health  University of California, San Francisco Medical Center  University of Colorado Hospital  University of Michigan Hospitals and Health Centers  UPMC Presbyterian Shadyside, Pittsburgh  View Large Table 2 Hospitals Responding in Full Barnes-Jewish Hospital/Washington University, St Louis  Brigham and Women’s Hospital  Cedars-Sinai Medical Center  Duke University Hospital  Hospitals of the University of Pennsylvania–Penn Presbyterian  Houston Methodist Hospital  Johns Hopkins Hospital  Massachusetts General Hospital  Mayo Clinic  Mount Sinai Hospital  New York-Presbyterian Columbia  New York-Presbyterian Cornell  Northwestern Memorial Hospital  Ronald Reagan UCLA Medical Center  Stanford Health Care–Stanford Hospital  The Cleveland Clinic  Tisch Hospital, NYU Langone Health  University of California, San Francisco Medical Center  University of Colorado Hospital  University of Michigan Hospitals and Health Centers  UPMC Presbyterian Shadyside, Pittsburgh  Barnes-Jewish Hospital/Washington University, St Louis  Brigham and Women’s Hospital  Cedars-Sinai Medical Center  Duke University Hospital  Hospitals of the University of Pennsylvania–Penn Presbyterian  Houston Methodist Hospital  Johns Hopkins Hospital  Massachusetts General Hospital  Mayo Clinic  Mount Sinai Hospital  New York-Presbyterian Columbia  New York-Presbyterian Cornell  Northwestern Memorial Hospital  Ronald Reagan UCLA Medical Center  Stanford Health Care–Stanford Hospital  The Cleveland Clinic  Tisch Hospital, NYU Langone Health  University of California, San Francisco Medical Center  University of Colorado Hospital  University of Michigan Hospitals and Health Centers  UPMC Presbyterian Shadyside, Pittsburgh  View Large Results Results were collected from all 21 institutions (100% response rate). Although some responses required additional communication to further clarify answers, eventually all questions were adequately addressed by all institutions. The instrument manufacturers used by the different institutions for routine CHEM and IM varied Table 3. For CHEM, only one institution used instruments from more than one vendor, whereas all other institutions used instruments from one vendor. The overall breakdown included Roche (Basel, Switzerland) (n = 10, 48%), Beckman Coulter (Brea, CA) (n = 6, 29%), Siemens (Berlin, Germany) (n = 4, 19%), and Abbott (Lake Bluff, IL) and Ortho (Raritan, NJ) (n = 1, 5% each) Table 3). In contrast, for IM, 10 (48%) of 21 institutions used instruments from multiple manufacturers (range, 1-5), including Roche (n = 13, 61.9%), Beckman Coulter (n = 10, 47.6%), Siemens (n = 9, 48.9%), Abbott (n = 7, 33.3%), Ortho (n = 1, 4.7%), and Bio-Rad (Hercules, CA) (n = 1, 4.7%). Table 3 Instruments Used for Automated Chemistry and Immunochemistry Hospital  Chemistry  Immunochemistry  A  Roche  Roche  B  Siemens  Siemens  C  Roche  Roche, Abbott  D  Siemens  Siemens  E  Roche  Roche  F  Roche, Beckman Coulter  Abbott, Roche, Beckman Coulter  G  Ortho  Ortho, Roche, Abbott  H  Beckman Coulter  Beckman Coulter, Roche  I  Beckman Coulter  Beckman Coulter, Siemens  J  Beckman Coulter  Roche, Abbott, Beckman Coulter, Siemens  K  Abbott  Abbott  L  Roche  Roche  M  Roche  Roche  N  Roche  Roche, Beckman Coulter, Siemens  O  Roche  Roche, Siemens, Beckman Coulter  P  Beckman Coulter  Beckman Coulter  Q  Roche  Roche, Abbott, Beckman Coulter  R  Siemens  Siemens  S  Siemens  Siemens  T  Roche  Roche, Siemens, Bio-Rad, Abbott, Beckman Coulter  U  Beckman Coulter  Beckman Coulter  Hospital  Chemistry  Immunochemistry  A  Roche  Roche  B  Siemens  Siemens  C  Roche  Roche, Abbott  D  Siemens  Siemens  E  Roche  Roche  F  Roche, Beckman Coulter  Abbott, Roche, Beckman Coulter  G  Ortho  Ortho, Roche, Abbott  H  Beckman Coulter  Beckman Coulter, Roche  I  Beckman Coulter  Beckman Coulter, Siemens  J  Beckman Coulter  Roche, Abbott, Beckman Coulter, Siemens  K  Abbott  Abbott  L  Roche  Roche  M  Roche  Roche  N  Roche  Roche, Beckman Coulter, Siemens  O  Roche  Roche, Siemens, Beckman Coulter  P  Beckman Coulter  Beckman Coulter  Q  Roche  Roche, Abbott, Beckman Coulter  R  Siemens  Siemens  S  Siemens  Siemens  T  Roche  Roche, Siemens, Bio-Rad, Abbott, Beckman Coulter  U  Beckman Coulter  Beckman Coulter  View Large Table 3 Instruments Used for Automated Chemistry and Immunochemistry Hospital  Chemistry  Immunochemistry  A  Roche  Roche  B  Siemens  Siemens  C  Roche  Roche, Abbott  D  Siemens  Siemens  E  Roche  Roche  F  Roche, Beckman Coulter  Abbott, Roche, Beckman Coulter  G  Ortho  Ortho, Roche, Abbott  H  Beckman Coulter  Beckman Coulter, Roche  I  Beckman Coulter  Beckman Coulter, Siemens  J  Beckman Coulter  Roche, Abbott, Beckman Coulter, Siemens  K  Abbott  Abbott  L  Roche  Roche  M  Roche  Roche  N  Roche  Roche, Beckman Coulter, Siemens  O  Roche  Roche, Siemens, Beckman Coulter  P  Beckman Coulter  Beckman Coulter  Q  Roche  Roche, Abbott, Beckman Coulter  R  Siemens  Siemens  S  Siemens  Siemens  T  Roche  Roche, Siemens, Bio-Rad, Abbott, Beckman Coulter  U  Beckman Coulter  Beckman Coulter  Hospital  Chemistry  Immunochemistry  A  Roche  Roche  B  Siemens  Siemens  C  Roche  Roche, Abbott  D  Siemens  Siemens  E  Roche  Roche  F  Roche, Beckman Coulter  Abbott, Roche, Beckman Coulter  G  Ortho  Ortho, Roche, Abbott  H  Beckman Coulter  Beckman Coulter, Roche  I  Beckman Coulter  Beckman Coulter, Siemens  J  Beckman Coulter  Roche, Abbott, Beckman Coulter, Siemens  K  Abbott  Abbott  L  Roche  Roche  M  Roche  Roche  N  Roche  Roche, Beckman Coulter, Siemens  O  Roche  Roche, Siemens, Beckman Coulter  P  Beckman Coulter  Beckman Coulter  Q  Roche  Roche, Abbott, Beckman Coulter  R  Siemens  Siemens  S  Siemens  Siemens  T  Roche  Roche, Siemens, Bio-Rad, Abbott, Beckman Coulter  U  Beckman Coulter  Beckman Coulter  View Large There was wide variation in the frequency of running CHEM QC Table 4, ranging from daily (n = 3, 14%) to every 2 hours (n = 2, 10%); intermediate intervals included every 4 hours (n = 3, 14%), 6 hours (n = 1, 5%), 8 hours (n = 6, 29%), and 12 hours (n = 6, 29%). Two (10%) institutions used a different QC frequency for electrolytes vs other CHEM tests. The total number of QC “events” per analyte (defined as the total number of times QC was run on an analyzer regardless of the number of levels) in a 24-hour period varied from a minimum of one to a maximum of 12. Three (14%) institutions used only one level per shift or one alternating level on the nonday shift hours (eg, a high control in the morning, a low control in the afternoon, and medium control on the late-night shifts). Table 4 Quality Control Frequency         QC Events/d  Hospital  CHEM  IM  Stat IM  CHEM  IM  Tn  hCG  A  2-3 Lv qd (electrolytes q8h)  2-3 Lv qd  2-3 Lv qd, negative QC q8h for Tn and hCG  3  1  3  3  B  2-3 Lv q8h  2-3 Lv q8h  2-3 Lv q8h  3  3  3  3  C  1 Lv (alternating) q2h  2-3 Lv q8h  cTn 4 Lv qd, 2 Lv alternating q2h, hCG 2 Lv q8h  12  3  12  3  D  2 Lv q12h  2 Lv q12h  2 Lv q8h  2  2  3  3  E  2 Lv q8h  2 Lv qd  2 Lv qd  3  1  1  1  F  2 Lv q8h  2 Lv qd  2 Lv qd  3  1  1  1  G  High control q12h, low control qd  2-3 Lv qd (medium Lv q12 for Lv 3 tests) or high Lv q12h/low Lv qd  Tn/CKMB/NT-proBNP: 2 Lv q12h; hCG: high q12h, low qd  2  2  2  2  H  2-3 Lv q12h  2-3 q12h  2-3 q12h  2  2  2  2  I  2-3 Lv q8h or qd  2-3 Lv qd  2-3 Lv qd  3  1  1  1  J  2 Lv q8h (q4h for electrolytes)  2-3 qd, 1 Lv (alternating) second/third shift  2-3 Lv day shift, 1 Lv (alternating) second/third shifts  6  3  3  3  K  2 Lv qd  2-3 Lv qd  Tn/hCG: 3 Lv qd  1  1  1  1  L  2 Lv q6h, 2 Lv startup/shutdown  2-3 Lv q12h  5 Lv q8h; hCG 2-3 Lv q12h  6  2  3  2  M  2 Lv at startup, 1 Lv (q4h alternating)  2-3 Lv at startup, 1 Lv at shutdown  Tn/hCG 2/3 Lv at startup, 1 Lv (alternating) q4h  7  2  7  7  N  3 Lv startup/shutdown, then 2 levels QC q2h  3 Lv startup/shutdown, plus 2 levels q8h  3 Lv q8h  12  4  3  3  O  3 Lv qd  2-3 Lv qd  Core laboratory: Tn/hCG 3 Lvq12h/q24 hours; ED laboratory: Tn/hCG 2 Lv qd  1  1  2  1  P  2-3 Lv q12h  2 Lv q8h  2 Lv q8h  2  3  3  3  Q  3 Lv day shift, 2 Lv other shifts  3 Lv day shift, 2 Lv other shifts  3 Lv qd, 2 Lv q12h  3  3  2  2  R  2 Lv q12h  2 Lv q12h  3 Lv q12h  2  2  2  2  S  2-3 q12h  q12h, certain tests once at start up  q12h, some qd  2  2  2  2  T  2 Lv qd  2 Lv qd  2 Lv qd  1  1  1  1  U  3 Lv q8h  3 Lv qd  3 Lv qd  3  1  1  1          QC Events/d  Hospital  CHEM  IM  Stat IM  CHEM  IM  Tn  hCG  A  2-3 Lv qd (electrolytes q8h)  2-3 Lv qd  2-3 Lv qd, negative QC q8h for Tn and hCG  3  1  3  3  B  2-3 Lv q8h  2-3 Lv q8h  2-3 Lv q8h  3  3  3  3  C  1 Lv (alternating) q2h  2-3 Lv q8h  cTn 4 Lv qd, 2 Lv alternating q2h, hCG 2 Lv q8h  12  3  12  3  D  2 Lv q12h  2 Lv q12h  2 Lv q8h  2  2  3  3  E  2 Lv q8h  2 Lv qd  2 Lv qd  3  1  1  1  F  2 Lv q8h  2 Lv qd  2 Lv qd  3  1  1  1  G  High control q12h, low control qd  2-3 Lv qd (medium Lv q12 for Lv 3 tests) or high Lv q12h/low Lv qd  Tn/CKMB/NT-proBNP: 2 Lv q12h; hCG: high q12h, low qd  2  2  2  2  H  2-3 Lv q12h  2-3 q12h  2-3 q12h  2  2  2  2  I  2-3 Lv q8h or qd  2-3 Lv qd  2-3 Lv qd  3  1  1  1  J  2 Lv q8h (q4h for electrolytes)  2-3 qd, 1 Lv (alternating) second/third shift  2-3 Lv day shift, 1 Lv (alternating) second/third shifts  6  3  3  3  K  2 Lv qd  2-3 Lv qd  Tn/hCG: 3 Lv qd  1  1  1  1  L  2 Lv q6h, 2 Lv startup/shutdown  2-3 Lv q12h  5 Lv q8h; hCG 2-3 Lv q12h  6  2  3  2  M  2 Lv at startup, 1 Lv (q4h alternating)  2-3 Lv at startup, 1 Lv at shutdown  Tn/hCG 2/3 Lv at startup, 1 Lv (alternating) q4h  7  2  7  7  N  3 Lv startup/shutdown, then 2 levels QC q2h  3 Lv startup/shutdown, plus 2 levels q8h  3 Lv q8h  12  4  3  3  O  3 Lv qd  2-3 Lv qd  Core laboratory: Tn/hCG 3 Lvq12h/q24 hours; ED laboratory: Tn/hCG 2 Lv qd  1  1  2  1  P  2-3 Lv q12h  2 Lv q8h  2 Lv q8h  2  3  3  3  Q  3 Lv day shift, 2 Lv other shifts  3 Lv day shift, 2 Lv other shifts  3 Lv qd, 2 Lv q12h  3  3  2  2  R  2 Lv q12h  2 Lv q12h  3 Lv q12h  2  2  2  2  S  2-3 q12h  q12h, certain tests once at start up  q12h, some qd  2  2  2  2  T  2 Lv qd  2 Lv qd  2 Lv qd  1  1  1  1  U  3 Lv q8h  3 Lv qd  3 Lv qd  3  1  1  1  CHEM, chemistry; CKMB, creatine kinase MB isoenzyme; cTn, cardiac troponin; ED, emergency department; hCG, human chorionic gonadotropin; IM, immunochemistry; Lv, level; NT-proBNP, NT-pro B-type natriuretic peptide; QC, quality control; QC Event, one or more QC levels run consecutively at a specified time on an analyzer; qd, per day; q(x)h, every x hours; Tn, troponin. View Large Table 4 Quality Control Frequency         QC Events/d  Hospital  CHEM  IM  Stat IM  CHEM  IM  Tn  hCG  A  2-3 Lv qd (electrolytes q8h)  2-3 Lv qd  2-3 Lv qd, negative QC q8h for Tn and hCG  3  1  3  3  B  2-3 Lv q8h  2-3 Lv q8h  2-3 Lv q8h  3  3  3  3  C  1 Lv (alternating) q2h  2-3 Lv q8h  cTn 4 Lv qd, 2 Lv alternating q2h, hCG 2 Lv q8h  12  3  12  3  D  2 Lv q12h  2 Lv q12h  2 Lv q8h  2  2  3  3  E  2 Lv q8h  2 Lv qd  2 Lv qd  3  1  1  1  F  2 Lv q8h  2 Lv qd  2 Lv qd  3  1  1  1  G  High control q12h, low control qd  2-3 Lv qd (medium Lv q12 for Lv 3 tests) or high Lv q12h/low Lv qd  Tn/CKMB/NT-proBNP: 2 Lv q12h; hCG: high q12h, low qd  2  2  2  2  H  2-3 Lv q12h  2-3 q12h  2-3 q12h  2  2  2  2  I  2-3 Lv q8h or qd  2-3 Lv qd  2-3 Lv qd  3  1  1  1  J  2 Lv q8h (q4h for electrolytes)  2-3 qd, 1 Lv (alternating) second/third shift  2-3 Lv day shift, 1 Lv (alternating) second/third shifts  6  3  3  3  K  2 Lv qd  2-3 Lv qd  Tn/hCG: 3 Lv qd  1  1  1  1  L  2 Lv q6h, 2 Lv startup/shutdown  2-3 Lv q12h  5 Lv q8h; hCG 2-3 Lv q12h  6  2  3  2  M  2 Lv at startup, 1 Lv (q4h alternating)  2-3 Lv at startup, 1 Lv at shutdown  Tn/hCG 2/3 Lv at startup, 1 Lv (alternating) q4h  7  2  7  7  N  3 Lv startup/shutdown, then 2 levels QC q2h  3 Lv startup/shutdown, plus 2 levels q8h  3 Lv q8h  12  4  3  3  O  3 Lv qd  2-3 Lv qd  Core laboratory: Tn/hCG 3 Lvq12h/q24 hours; ED laboratory: Tn/hCG 2 Lv qd  1  1  2  1  P  2-3 Lv q12h  2 Lv q8h  2 Lv q8h  2  3  3  3  Q  3 Lv day shift, 2 Lv other shifts  3 Lv day shift, 2 Lv other shifts  3 Lv qd, 2 Lv q12h  3  3  2  2  R  2 Lv q12h  2 Lv q12h  3 Lv q12h  2  2  2  2  S  2-3 q12h  q12h, certain tests once at start up  q12h, some qd  2  2  2  2  T  2 Lv qd  2 Lv qd  2 Lv qd  1  1  1  1  U  3 Lv q8h  3 Lv qd  3 Lv qd  3  1  1  1          QC Events/d  Hospital  CHEM  IM  Stat IM  CHEM  IM  Tn  hCG  A  2-3 Lv qd (electrolytes q8h)  2-3 Lv qd  2-3 Lv qd, negative QC q8h for Tn and hCG  3  1  3  3  B  2-3 Lv q8h  2-3 Lv q8h  2-3 Lv q8h  3  3  3  3  C  1 Lv (alternating) q2h  2-3 Lv q8h  cTn 4 Lv qd, 2 Lv alternating q2h, hCG 2 Lv q8h  12  3  12  3  D  2 Lv q12h  2 Lv q12h  2 Lv q8h  2  2  3  3  E  2 Lv q8h  2 Lv qd  2 Lv qd  3  1  1  1  F  2 Lv q8h  2 Lv qd  2 Lv qd  3  1  1  1  G  High control q12h, low control qd  2-3 Lv qd (medium Lv q12 for Lv 3 tests) or high Lv q12h/low Lv qd  Tn/CKMB/NT-proBNP: 2 Lv q12h; hCG: high q12h, low qd  2  2  2  2  H  2-3 Lv q12h  2-3 q12h  2-3 q12h  2  2  2  2  I  2-3 Lv q8h or qd  2-3 Lv qd  2-3 Lv qd  3  1  1  1  J  2 Lv q8h (q4h for electrolytes)  2-3 qd, 1 Lv (alternating) second/third shift  2-3 Lv day shift, 1 Lv (alternating) second/third shifts  6  3  3  3  K  2 Lv qd  2-3 Lv qd  Tn/hCG: 3 Lv qd  1  1  1  1  L  2 Lv q6h, 2 Lv startup/shutdown  2-3 Lv q12h  5 Lv q8h; hCG 2-3 Lv q12h  6  2  3  2  M  2 Lv at startup, 1 Lv (q4h alternating)  2-3 Lv at startup, 1 Lv at shutdown  Tn/hCG 2/3 Lv at startup, 1 Lv (alternating) q4h  7  2  7  7  N  3 Lv startup/shutdown, then 2 levels QC q2h  3 Lv startup/shutdown, plus 2 levels q8h  3 Lv q8h  12  4  3  3  O  3 Lv qd  2-3 Lv qd  Core laboratory: Tn/hCG 3 Lvq12h/q24 hours; ED laboratory: Tn/hCG 2 Lv qd  1  1  2  1  P  2-3 Lv q12h  2 Lv q8h  2 Lv q8h  2  3  3  3  Q  3 Lv day shift, 2 Lv other shifts  3 Lv day shift, 2 Lv other shifts  3 Lv qd, 2 Lv q12h  3  3  2  2  R  2 Lv q12h  2 Lv q12h  3 Lv q12h  2  2  2  2  S  2-3 q12h  q12h, certain tests once at start up  q12h, some qd  2  2  2  2  T  2 Lv qd  2 Lv qd  2 Lv qd  1  1  1  1  U  3 Lv q8h  3 Lv qd  3 Lv qd  3  1  1  1  CHEM, chemistry; CKMB, creatine kinase MB isoenzyme; cTn, cardiac troponin; ED, emergency department; hCG, human chorionic gonadotropin; IM, immunochemistry; Lv, level; NT-proBNP, NT-pro B-type natriuretic peptide; QC, quality control; QC Event, one or more QC levels run consecutively at a specified time on an analyzer; qd, per day; q(x)h, every x hours; Tn, troponin. View Large There was also variability in the frequency and number of QC for IM and STAT IM (eg, troponin and β–human chorionic gonadotropin [hCG]). The total number of non-STAT IM QC levels used per day varied between institutions from a minimum of two levels to a maximum of three different levels per day (Table 4). QC events per 24 hours were less varied than for CHEMs, ranging from one to seven. For STAT IMs, 19 laboratories used two or three levels, with one laboratory using four levels for troponin (which included a QC near the 99% cutoff), another laboratory using five levels for troponin, and four laboratories using alternating levels at different time points. QC events for STAT IM per 24 hours varied from one to 12 per day. As with CHEM, some hospitals used a single control level on select shifts. For CHEM, many hospitals used QC materials from multiple sources (n = 8, 38%) Table 5. However, all respondents predominantly relied on third-party QC materials (n = 21, 100%). Of the third-party reagents, Bio-Rad was the most common (n = 19, 90%), with MAS ChemTrak (Thermo Fisher Scientific, Waltham, MA) less commonly used (n = 2, 10%). For IM, 12 (57%) laboratories used more than one source for control materials. Again, Bio-Rad QC materials were the most commonly used (n = 19, 90%), and two (10%) laboratories favored the use of manufacturer QC materials over third party. Table 5 Quality Control Material Vendors Hospital  Chemistry  Immunochemistry  A  MAS Chemtrak  Bio-Rad  B  Bio-Rad  Bio-Rad  C  Bio-Rad  Bio-Rad  D  Bio-Rad  Bio-Rad  E  Bio-Rad, manufacturer when necessary  Bio-Rad, manufacturer QC when necessary  F  Bio-Rad  Bio-Rad  G  Bio-Rad  Bio-Rad  H  Bio-Rad  Bio-Rad  I  Predominantly Bio- Rad QC and some manufacturer supplied (if available)  Predominantly Bio-Rad QC and some manufacturer supplied (if available)  J  Bio-Rad, MAS Chemtrak  Automatic laboratory: manufacturer supplied when possible, third party when necessary (Bio-Rad). Special chemistry laboratory: Bio-Rad when possible, manufacturer when necessary; supplemented with patient pools  K  Bio-Rad  Bio-Rad, manufacturer QC when necessary  L  Bio-Rad  Bio-Rad, Roche, Randox  M  MAS ChemTrack  Manufacturer, third party when needed (Bio-Rad)  N  Mostly Bio-Rad, also Stanbio, Thermo TDM, UTAK, Elsohly  Mostly Bio-Rad, manufacturer control where necessary  O  Bio-Rad, very few manufacturers or laboratory made  Bio-Rad, very few manufacturers  P  Bio-Rad  Bio-Rad  Q  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  R  Bio-Rad  Bio-Rad  S  Bio-Rad  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  T  Bio-Rad, some Roche  Bio-Rad, some Roche, Abbott  U  Bio-Rad, very few manufacturers  Bio-Rad, very few manufacturers  Hospital  Chemistry  Immunochemistry  A  MAS Chemtrak  Bio-Rad  B  Bio-Rad  Bio-Rad  C  Bio-Rad  Bio-Rad  D  Bio-Rad  Bio-Rad  E  Bio-Rad, manufacturer when necessary  Bio-Rad, manufacturer QC when necessary  F  Bio-Rad  Bio-Rad  G  Bio-Rad  Bio-Rad  H  Bio-Rad  Bio-Rad  I  Predominantly Bio- Rad QC and some manufacturer supplied (if available)  Predominantly Bio-Rad QC and some manufacturer supplied (if available)  J  Bio-Rad, MAS Chemtrak  Automatic laboratory: manufacturer supplied when possible, third party when necessary (Bio-Rad). Special chemistry laboratory: Bio-Rad when possible, manufacturer when necessary; supplemented with patient pools  K  Bio-Rad  Bio-Rad, manufacturer QC when necessary  L  Bio-Rad  Bio-Rad, Roche, Randox  M  MAS ChemTrack  Manufacturer, third party when needed (Bio-Rad)  N  Mostly Bio-Rad, also Stanbio, Thermo TDM, UTAK, Elsohly  Mostly Bio-Rad, manufacturer control where necessary  O  Bio-Rad, very few manufacturers or laboratory made  Bio-Rad, very few manufacturers  P  Bio-Rad  Bio-Rad  Q  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  R  Bio-Rad  Bio-Rad  S  Bio-Rad  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  T  Bio-Rad, some Roche  Bio-Rad, some Roche, Abbott  U  Bio-Rad, very few manufacturers  Bio-Rad, very few manufacturers  QC, quality control. View Large Table 5 Quality Control Material Vendors Hospital  Chemistry  Immunochemistry  A  MAS Chemtrak  Bio-Rad  B  Bio-Rad  Bio-Rad  C  Bio-Rad  Bio-Rad  D  Bio-Rad  Bio-Rad  E  Bio-Rad, manufacturer when necessary  Bio-Rad, manufacturer QC when necessary  F  Bio-Rad  Bio-Rad  G  Bio-Rad  Bio-Rad  H  Bio-Rad  Bio-Rad  I  Predominantly Bio- Rad QC and some manufacturer supplied (if available)  Predominantly Bio-Rad QC and some manufacturer supplied (if available)  J  Bio-Rad, MAS Chemtrak  Automatic laboratory: manufacturer supplied when possible, third party when necessary (Bio-Rad). Special chemistry laboratory: Bio-Rad when possible, manufacturer when necessary; supplemented with patient pools  K  Bio-Rad  Bio-Rad, manufacturer QC when necessary  L  Bio-Rad  Bio-Rad, Roche, Randox  M  MAS ChemTrack  Manufacturer, third party when needed (Bio-Rad)  N  Mostly Bio-Rad, also Stanbio, Thermo TDM, UTAK, Elsohly  Mostly Bio-Rad, manufacturer control where necessary  O  Bio-Rad, very few manufacturers or laboratory made  Bio-Rad, very few manufacturers  P  Bio-Rad  Bio-Rad  Q  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  R  Bio-Rad  Bio-Rad  S  Bio-Rad  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  T  Bio-Rad, some Roche  Bio-Rad, some Roche, Abbott  U  Bio-Rad, very few manufacturers  Bio-Rad, very few manufacturers  Hospital  Chemistry  Immunochemistry  A  MAS Chemtrak  Bio-Rad  B  Bio-Rad  Bio-Rad  C  Bio-Rad  Bio-Rad  D  Bio-Rad  Bio-Rad  E  Bio-Rad, manufacturer when necessary  Bio-Rad, manufacturer QC when necessary  F  Bio-Rad  Bio-Rad  G  Bio-Rad  Bio-Rad  H  Bio-Rad  Bio-Rad  I  Predominantly Bio- Rad QC and some manufacturer supplied (if available)  Predominantly Bio-Rad QC and some manufacturer supplied (if available)  J  Bio-Rad, MAS Chemtrak  Automatic laboratory: manufacturer supplied when possible, third party when necessary (Bio-Rad). Special chemistry laboratory: Bio-Rad when possible, manufacturer when necessary; supplemented with patient pools  K  Bio-Rad  Bio-Rad, manufacturer QC when necessary  L  Bio-Rad  Bio-Rad, Roche, Randox  M  MAS ChemTrack  Manufacturer, third party when needed (Bio-Rad)  N  Mostly Bio-Rad, also Stanbio, Thermo TDM, UTAK, Elsohly  Mostly Bio-Rad, manufacturer control where necessary  O  Bio-Rad, very few manufacturers or laboratory made  Bio-Rad, very few manufacturers  P  Bio-Rad  Bio-Rad  Q  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  R  Bio-Rad  Bio-Rad  S  Bio-Rad  Bio-Rad, some Beckman Coulter, Abbott, and in-house prepared  T  Bio-Rad, some Roche  Bio-Rad, some Roche, Abbott  U  Bio-Rad, very few manufacturers  Bio-Rad, very few manufacturers  QC, quality control. View Large Most hospitals used a QC rule of 2 SD (n = 16, 76%), two (10%) used variable rules (based on the test) between 2 and 3 SD, and one (5%) used a cutoff of 3 SD Table 6. Two (10%) hospitals used derivations of the Westgard rules depending on the assay. For IM, 17 (81%) hospitals chose a 2-SD rule, one (5%) chose 3 SD, and three (14%) chose some permutation of the Westgard rules. Table 6 Quality Control Rules Hospital  Chemistry  Immunochemistry  A  ±2 SD  ±2 SD  B  ±2 SD  ±2 SD  C  ±2 SD  ±2 SD  D  ±2 SD, some ±2.5 SD and ±3 SD  ±2 SD  E  ±2 SD  ±2 SD  F  ±2 SD  ±2 SD  G  ±2 SD  ±2 SD  H  ±2 SD  ±2 SD  I  Variable (1-3S, 1-3.5S, 1-4S, 1-5S, 2-2S, 2 of 3-2S, R-4S, 3-1S, 4-1S, or N-x, 1-2S, 1-2.5S)  Variable (1-3S, 1-3.5S, 1-4S, 1-5S, 2-2S, 2 of 3-2S, R-4S, 3-1S, 4-1S, or N-x, 1-2S, 1-2.5S)  J  ±2 SD  ±2 SD  K  ±2 SD  ±2 SD  L  ±2 SD  ±2 SD  M  ±2 SD  ±2 SD  N  ±2 SD  ±2 SD  O  ±2 SD  ±2 SD  P  ±2 SD  ±2 SD  Q  3 SD  3 SD  R  ±2 SD  ±2 SD  S  2-2S, 1-3S, R-4S, 10x  2-2S, 1-3S, R-4S, 10x  T  2 or 3 SD  Some 2 or 3 SD; others Westgard rules  U  ±2 SD  ±2 SD  Hospital  Chemistry  Immunochemistry  A  ±2 SD  ±2 SD  B  ±2 SD  ±2 SD  C  ±2 SD  ±2 SD  D  ±2 SD, some ±2.5 SD and ±3 SD  ±2 SD  E  ±2 SD  ±2 SD  F  ±2 SD  ±2 SD  G  ±2 SD  ±2 SD  H  ±2 SD  ±2 SD  I  Variable (1-3S, 1-3.5S, 1-4S, 1-5S, 2-2S, 2 of 3-2S, R-4S, 3-1S, 4-1S, or N-x, 1-2S, 1-2.5S)  Variable (1-3S, 1-3.5S, 1-4S, 1-5S, 2-2S, 2 of 3-2S, R-4S, 3-1S, 4-1S, or N-x, 1-2S, 1-2.5S)  J  ±2 SD  ±2 SD  K  ±2 SD  ±2 SD  L  ±2 SD  ±2 SD  M  ±2 SD  ±2 SD  N  ±2 SD  ±2 SD  O  ±2 SD  ±2 SD  P  ±2 SD  ±2 SD  Q  3 SD  3 SD  R  ±2 SD  ±2 SD  S  2-2S, 1-3S, R-4S, 10x  2-2S, 1-3S, R-4S, 10x  T  2 or 3 SD  Some 2 or 3 SD; others Westgard rules  U  ±2 SD  ±2 SD  View Large Table 6 Quality Control Rules Hospital  Chemistry  Immunochemistry  A  ±2 SD  ±2 SD  B  ±2 SD  ±2 SD  C  ±2 SD  ±2 SD  D  ±2 SD, some ±2.5 SD and ±3 SD  ±2 SD  E  ±2 SD  ±2 SD  F  ±2 SD  ±2 SD  G  ±2 SD  ±2 SD  H  ±2 SD  ±2 SD  I  Variable (1-3S, 1-3.5S, 1-4S, 1-5S, 2-2S, 2 of 3-2S, R-4S, 3-1S, 4-1S, or N-x, 1-2S, 1-2.5S)  Variable (1-3S, 1-3.5S, 1-4S, 1-5S, 2-2S, 2 of 3-2S, R-4S, 3-1S, 4-1S, or N-x, 1-2S, 1-2.5S)  J  ±2 SD  ±2 SD  K  ±2 SD  ±2 SD  L  ±2 SD  ±2 SD  M  ±2 SD  ±2 SD  N  ±2 SD  ±2 SD  O  ±2 SD  ±2 SD  P  ±2 SD  ±2 SD  Q  3 SD  3 SD  R  ±2 SD  ±2 SD  S  2-2S, 1-3S, R-4S, 10x  2-2S, 1-3S, R-4S, 10x  T  2 or 3 SD  Some 2 or 3 SD; others Westgard rules  U  ±2 SD  ±2 SD  Hospital  Chemistry  Immunochemistry  A  ±2 SD  ±2 SD  B  ±2 SD  ±2 SD  C  ±2 SD  ±2 SD  D  ±2 SD, some ±2.5 SD and ±3 SD  ±2 SD  E  ±2 SD  ±2 SD  F  ±2 SD  ±2 SD  G  ±2 SD  ±2 SD  H  ±2 SD  ±2 SD  I  Variable (1-3S, 1-3.5S, 1-4S, 1-5S, 2-2S, 2 of 3-2S, R-4S, 3-1S, 4-1S, or N-x, 1-2S, 1-2.5S)  Variable (1-3S, 1-3.5S, 1-4S, 1-5S, 2-2S, 2 of 3-2S, R-4S, 3-1S, 4-1S, or N-x, 1-2S, 1-2.5S)  J  ±2 SD  ±2 SD  K  ±2 SD  ±2 SD  L  ±2 SD  ±2 SD  M  ±2 SD  ±2 SD  N  ±2 SD  ±2 SD  O  ±2 SD  ±2 SD  P  ±2 SD  ±2 SD  Q  3 SD  3 SD  R  ±2 SD  ±2 SD  S  2-2S, 1-3S, R-4S, 10x  2-2S, 1-3S, R-4S, 10x  T  2 or 3 SD  Some 2 or 3 SD; others Westgard rules  U  ±2 SD  ±2 SD  View Large When a QC was out of control, all but one hospital elected to repeat the control and accept results if the repeat came back into control (n = 20, 95%), although some had some minor variations (such as rejecting a run outright if a control was out by >4 SD; Table 7). One (5%) institution rejected runs if two controls were out 2 SD or if one was out 3 SD. Another institution (5%) repeated QC if one of the two levels was out 2 SD from target value but accepted the run if only one of three QC levels was out. This institution also rejected outright if the following rules were broken: 1- 2.5S, 2-2S, 2/3-2S, or R-4S. Only laboratory E made no provisions for repeating control(s); if one level exceeded 3 SD or two controls exceeded 2 SD, the run was rejected as being out of control. Table 7 Quality Control Rules Hospital  QC Rules  A  If control is out of range, it is repeated. If repeat is in range, the results are accepted. If still out of range, assay is recalibrated and/or additional troubleshooting occurs (eg, new reagent pack, new QC). Patient results may be repeated if it is determined that QC was out when patient results were reported.  B  As “A.”  C  As “A.”  D  As “A.”  E  Reject results if two QC values are out 2 SD or more or if one is out 3 SD. If QC fails, corrective action is taken. If one QC is out 2 SD, evaluate other QC in same run and in previous runs (warning only).  F  As “A.”  G  As “A.”  H  As “A” if one QC out <4 SD. If two QCs are out 2 to 4 SD, both must be repeated and corrective action is taken if one or more are still out. If QC out >4 SD, stop analysis immediately and take corrective action before continuing testing.  I  As “A.”  J  As “A”  K  As “A,” but if the QC results are acceptable after recalibration, 10 specimens analyzed within the last 24 hours are repeated. If they are within 10% of reported value, the assay is validated. If not, further corrective action is taken.  L  As “A.”  M  As “A.”  N  As “A.”  O  For analytes with two QC levels and troponin, as “A” but also rejected if 1-2.5S, 2-2S, 2/3-2S, or R-4S rules are violated. We also have a warning with the 7-T rule. For analytes with three QC levels, if control is outside of 2 SD range (1-2S), it is a warning; repeat is not required and run is accepted. If two QC levels are outside of 2 SD range (2-2S), run is rejected, and controls are repeated using new vials/aliquots of controls. If QC is back into range, the run is accepted. If repeat is out of range, investigate/recalibrate. Reject run also if 1-2.5S, 2-2S, 2/3-2S, or R-4S rule is violated. We also have a warning with the 7-T rule.  P  As “A.”  Q  If QC is out 3 SD, then QC failure; repeat once, and if it fails again, take corrective action.  R  We use several Westgard rules (4-1S, 10x) to further investigate when a 2-SD flag occurs on QC. First step is to repeat QC; if it is in and no Westgard rules have triggered, accept and continue. If out, check QC on other analyzers and rerun at least five patients on a different analyzer. Notify supervisor if recalibration is warranted.  S  If results are out by 2-2S or 1-3S, the QC is repeated and accepted if it comes in. If it is out, the system is recalibrated and QC is rerun. Any further problems are escalated to the lead in the area for further troubleshooting.  T  As “A.”  U  As “A”; if far outside of 2 SD, then may be worked up at coordinator’s discretion.  Hospital  QC Rules  A  If control is out of range, it is repeated. If repeat is in range, the results are accepted. If still out of range, assay is recalibrated and/or additional troubleshooting occurs (eg, new reagent pack, new QC). Patient results may be repeated if it is determined that QC was out when patient results were reported.  B  As “A.”  C  As “A.”  D  As “A.”  E  Reject results if two QC values are out 2 SD or more or if one is out 3 SD. If QC fails, corrective action is taken. If one QC is out 2 SD, evaluate other QC in same run and in previous runs (warning only).  F  As “A.”  G  As “A.”  H  As “A” if one QC out <4 SD. If two QCs are out 2 to 4 SD, both must be repeated and corrective action is taken if one or more are still out. If QC out >4 SD, stop analysis immediately and take corrective action before continuing testing.  I  As “A.”  J  As “A”  K  As “A,” but if the QC results are acceptable after recalibration, 10 specimens analyzed within the last 24 hours are repeated. If they are within 10% of reported value, the assay is validated. If not, further corrective action is taken.  L  As “A.”  M  As “A.”  N  As “A.”  O  For analytes with two QC levels and troponin, as “A” but also rejected if 1-2.5S, 2-2S, 2/3-2S, or R-4S rules are violated. We also have a warning with the 7-T rule. For analytes with three QC levels, if control is outside of 2 SD range (1-2S), it is a warning; repeat is not required and run is accepted. If two QC levels are outside of 2 SD range (2-2S), run is rejected, and controls are repeated using new vials/aliquots of controls. If QC is back into range, the run is accepted. If repeat is out of range, investigate/recalibrate. Reject run also if 1-2.5S, 2-2S, 2/3-2S, or R-4S rule is violated. We also have a warning with the 7-T rule.  P  As “A.”  Q  If QC is out 3 SD, then QC failure; repeat once, and if it fails again, take corrective action.  R  We use several Westgard rules (4-1S, 10x) to further investigate when a 2-SD flag occurs on QC. First step is to repeat QC; if it is in and no Westgard rules have triggered, accept and continue. If out, check QC on other analyzers and rerun at least five patients on a different analyzer. Notify supervisor if recalibration is warranted.  S  If results are out by 2-2S or 1-3S, the QC is repeated and accepted if it comes in. If it is out, the system is recalibrated and QC is rerun. Any further problems are escalated to the lead in the area for further troubleshooting.  T  As “A.”  U  As “A”; if far outside of 2 SD, then may be worked up at coordinator’s discretion.  QC, quality control. View Large Table 7 Quality Control Rules Hospital  QC Rules  A  If control is out of range, it is repeated. If repeat is in range, the results are accepted. If still out of range, assay is recalibrated and/or additional troubleshooting occurs (eg, new reagent pack, new QC). Patient results may be repeated if it is determined that QC was out when patient results were reported.  B  As “A.”  C  As “A.”  D  As “A.”  E  Reject results if two QC values are out 2 SD or more or if one is out 3 SD. If QC fails, corrective action is taken. If one QC is out 2 SD, evaluate other QC in same run and in previous runs (warning only).  F  As “A.”  G  As “A.”  H  As “A” if one QC out <4 SD. If two QCs are out 2 to 4 SD, both must be repeated and corrective action is taken if one or more are still out. If QC out >4 SD, stop analysis immediately and take corrective action before continuing testing.  I  As “A.”  J  As “A”  K  As “A,” but if the QC results are acceptable after recalibration, 10 specimens analyzed within the last 24 hours are repeated. If they are within 10% of reported value, the assay is validated. If not, further corrective action is taken.  L  As “A.”  M  As “A.”  N  As “A.”  O  For analytes with two QC levels and troponin, as “A” but also rejected if 1-2.5S, 2-2S, 2/3-2S, or R-4S rules are violated. We also have a warning with the 7-T rule. For analytes with three QC levels, if control is outside of 2 SD range (1-2S), it is a warning; repeat is not required and run is accepted. If two QC levels are outside of 2 SD range (2-2S), run is rejected, and controls are repeated using new vials/aliquots of controls. If QC is back into range, the run is accepted. If repeat is out of range, investigate/recalibrate. Reject run also if 1-2.5S, 2-2S, 2/3-2S, or R-4S rule is violated. We also have a warning with the 7-T rule.  P  As “A.”  Q  If QC is out 3 SD, then QC failure; repeat once, and if it fails again, take corrective action.  R  We use several Westgard rules (4-1S, 10x) to further investigate when a 2-SD flag occurs on QC. First step is to repeat QC; if it is in and no Westgard rules have triggered, accept and continue. If out, check QC on other analyzers and rerun at least five patients on a different analyzer. Notify supervisor if recalibration is warranted.  S  If results are out by 2-2S or 1-3S, the QC is repeated and accepted if it comes in. If it is out, the system is recalibrated and QC is rerun. Any further problems are escalated to the lead in the area for further troubleshooting.  T  As “A.”  U  As “A”; if far outside of 2 SD, then may be worked up at coordinator’s discretion.  Hospital  QC Rules  A  If control is out of range, it is repeated. If repeat is in range, the results are accepted. If still out of range, assay is recalibrated and/or additional troubleshooting occurs (eg, new reagent pack, new QC). Patient results may be repeated if it is determined that QC was out when patient results were reported.  B  As “A.”  C  As “A.”  D  As “A.”  E  Reject results if two QC values are out 2 SD or more or if one is out 3 SD. If QC fails, corrective action is taken. If one QC is out 2 SD, evaluate other QC in same run and in previous runs (warning only).  F  As “A.”  G  As “A.”  H  As “A” if one QC out <4 SD. If two QCs are out 2 to 4 SD, both must be repeated and corrective action is taken if one or more are still out. If QC out >4 SD, stop analysis immediately and take corrective action before continuing testing.  I  As “A.”  J  As “A”  K  As “A,” but if the QC results are acceptable after recalibration, 10 specimens analyzed within the last 24 hours are repeated. If they are within 10% of reported value, the assay is validated. If not, further corrective action is taken.  L  As “A.”  M  As “A.”  N  As “A.”  O  For analytes with two QC levels and troponin, as “A” but also rejected if 1-2.5S, 2-2S, 2/3-2S, or R-4S rules are violated. We also have a warning with the 7-T rule. For analytes with three QC levels, if control is outside of 2 SD range (1-2S), it is a warning; repeat is not required and run is accepted. If two QC levels are outside of 2 SD range (2-2S), run is rejected, and controls are repeated using new vials/aliquots of controls. If QC is back into range, the run is accepted. If repeat is out of range, investigate/recalibrate. Reject run also if 1-2.5S, 2-2S, 2/3-2S, or R-4S rule is violated. We also have a warning with the 7-T rule.  P  As “A.”  Q  If QC is out 3 SD, then QC failure; repeat once, and if it fails again, take corrective action.  R  We use several Westgard rules (4-1S, 10x) to further investigate when a 2-SD flag occurs on QC. First step is to repeat QC; if it is in and no Westgard rules have triggered, accept and continue. If out, check QC on other analyzers and rerun at least five patients on a different analyzer. Notify supervisor if recalibration is warranted.  S  If results are out by 2-2S or 1-3S, the QC is repeated and accepted if it comes in. If it is out, the system is recalibrated and QC is rerun. Any further problems are escalated to the lead in the area for further troubleshooting.  T  As “A.”  U  As “A”; if far outside of 2 SD, then may be worked up at coordinator’s discretion.  QC, quality control. View Large Although most of the surveyed hospitals do not currently use moving averages (n = 19, 90%), four (19%) are hoping to implement moving averages in the near future Table 8. One (5%) institution runs moving averages in the background but does not use them for clinical metrics. Only one (5%) uses moving averages for clinical use and then only for a small number of assays. One of the institutions that did not use moving averages reported that it had previously implemented them but had not found them to be useful. Table 8 Moving Averages Hospital  Do You Use Moving Averages?  A  No  B  No  C  No  D  No  E  No  F  Not yet, but planning to  G  No  H  No  I  No  J  No [but hopefully soon]  K  No  L  Not clinically, run for some in background to collect data  M  No  N  No, investigated but not useful  O  No  P  No  Q  No  R  We do use moving averages routinely as QC point on six different assays and have the capability of turning it on for other assays if deemed necessary.  S  We are collecting data currently and will be implementing spring/summer of 2017.  T  No  U  Not at this time, but planning to implement this year  Hospital  Do You Use Moving Averages?  A  No  B  No  C  No  D  No  E  No  F  Not yet, but planning to  G  No  H  No  I  No  J  No [but hopefully soon]  K  No  L  Not clinically, run for some in background to collect data  M  No  N  No, investigated but not useful  O  No  P  No  Q  No  R  We do use moving averages routinely as QC point on six different assays and have the capability of turning it on for other assays if deemed necessary.  S  We are collecting data currently and will be implementing spring/summer of 2017.  T  No  U  Not at this time, but planning to implement this year  QC, quality control. View Large Table 8 Moving Averages Hospital  Do You Use Moving Averages?  A  No  B  No  C  No  D  No  E  No  F  Not yet, but planning to  G  No  H  No  I  No  J  No [but hopefully soon]  K  No  L  Not clinically, run for some in background to collect data  M  No  N  No, investigated but not useful  O  No  P  No  Q  No  R  We do use moving averages routinely as QC point on six different assays and have the capability of turning it on for other assays if deemed necessary.  S  We are collecting data currently and will be implementing spring/summer of 2017.  T  No  U  Not at this time, but planning to implement this year  Hospital  Do You Use Moving Averages?  A  No  B  No  C  No  D  No  E  No  F  Not yet, but planning to  G  No  H  No  I  No  J  No [but hopefully soon]  K  No  L  Not clinically, run for some in background to collect data  M  No  N  No, investigated but not useful  O  No  P  No  Q  No  R  We do use moving averages routinely as QC point on six different assays and have the capability of turning it on for other assays if deemed necessary.  S  We are collecting data currently and will be implementing spring/summer of 2017.  T  No  U  Not at this time, but planning to implement this year  QC, quality control. View Large Discussion QC is a critical aspect of laboratory management that laboratory directors take very seriously as it helps ensure we provide accurate results to guide clinical management. This is perhaps related to the high rate of response (100%) from our cohort. In this study, we were able to survey a wide range of highly regarded academic institutions comprising the entirety of the US News & World Report 2016 to 2017 honor roll list.7 This list consisted of 21 hospitals from 12 states. We selected this cohort as a group of high-performing academic institutions with diverse practice settings and distinct academic histories. As expected, a variety of different instruments were used in the laboratories surveyed. Although some vendors were seen more frequently among respondents, no single vendor or platform had complete market dominance. Since QC rules are often related to actual assay performance, it makes sense that different platforms, which will have distinct methods, will perform somewhat differently and may require different QC rules. There were no apparent trends between QC frequency and manufacturer, and variation existed between laboratories using similar instruments. Therefore, it seems unlikely that platform choice alone would affect QC practices to a significant extent. A factor that supports this is that, in most institutions that used instruments from multiple manufacturers, QC rules did not vary notably between the instruments. There was dramatic (ie, 12-fold) variation in QC frequency, ranging from once daily to every 2 hours. This was surprising because, although QC frequency may vary based on the device used, reagent stability, and test volume, the clinical risk associated with result errors should be more or less similar among the cohort, especially for routine CHEM/IM testing. Possible factors include assay method, test volume, economic constraints, and the difficulty of repeating patient samples in the event of QC failure. When QC fails, there is the possibility of needing to repeat all patient samples tested since the previous successful QC. For a high-volume test, potentially repeating all samples from the previous 24 hours would result in a very large number of samples that must be retrieved and retested. In addition, a significant delay before erroneous results are corrected has the potential to affect patient care. These are reasonable concerns, but increased frequency of QC may result in increased QC failures for purely statistical reasons, resulting in unnecessary corrective actions, delayed results, and increased costs. Other QC practices were reported by several laboratories in the survey. At the time of surveying, three laboratories (C, J, and M) performed “alternating-level” QC testing—that is, instead of testing two levels of QC material at t = X hours after daily startup, one level was tested at t = X/2 hours and an alternate level tested at t = X hours. Laboratories C and M performed such “alternating” QC testing for CHEMs, all three for STAT IM (laboratory C did for troponin but not for hCG), and only laboratory J for IMs (in the time since surveying, laboratory J has ceased performing alternating-level QC). In theory, this practice reduces the time to detect an out-of-control situation compared with performing both QC levels at double the time interval. This practice assumes that all out-of-control situations will be detected equally well by testing either QC level, an assumption that may not apply in all analytical circumstances. Several laboratories reduced the number of QC levels tested after daily startup; for example, if three levels of QC were tested to demonstrate an instrument test was “in control” at time t = 0 hours, subsequent QC testing events might only employ one or two QC levels over the subsequent 24 hours. Four laboratories (G, M, N, and Q) did this for CHEM, five for IM (G, J, M, N, and Q), and six for STAT IM (A, C [troponin only], G [hCG only], J, M, and Q). Two laboratories (A and J) performed QC testing for electrolytes (Na, K, and Cl) at a higher frequency than CHEM, likely due to the high volume of tests. QC materials were overwhelmingly third party, with only a single laboratory relying predominantly on manufacturer-supplied QC materials (for IM only). Third-party materials have the theoretical advantage of providing a more independent verification of assay function and may have the option of QC material return to the parent company for further analysis if repeated failure occurs with a lot. On the other hand, manufacturer-produced materials have the potential benefit of being specifically designed for the system and test in question such that assay failures can be traced to a single manufacturer source, as opposed to having to query two separate vendors for QC material and machine issues. Despite these theoretical considerations, there has not been a systematic study into which method provides the most reliable approach for the selection of QC materials. Most (n = 16, 76%) respondents used a QC range of 2 SD almost exclusively, and 14% (n = 3) used a combination of between 2 and 3 SD. This is an unexpected finding, as there is no standard Westgard rule for 1-2S, except as a warning. Assuming a normal distribution of values for QC materials on repeat analysis (ie, normal random variation and no systemic bias), a QC will be out of a 2-SD control range approximately 5% of the time compared with only 1% if a cutoff of 3 SD is used. Although a 2-SD QC rule has an increased chance of finding small analytical variations, one would expect lowered specificity, with numerous incidences of QC out-of-range results merely due to chance. In contrast, in laboratories where 3 SD or 2 × 2 SD are used, one would theoretically expect a potential 1% or 0.25% rate of QC out of control based on random error. Of the 21 hospitals that responded to the survey, only two explicitly used Westgard rule derivations. These rules were introduced in the 1970s by James Westgard and colleagues in an effort to apply a mathematically rigorous approach to systematizing quality control. The Westgard rules are used to evaluate QC data and are designed to capture both increased variation (random error) and bias (systemic error) while minimizing both false negatives and false positives.8 Minimization of false negatives is especially critical, since these patients may be entirely missed or discharged without proper workup (as opposed to false positives, which may be revelated by further testing). There are several iterations of the Westgard rules, but the process of flagging and repeating controls out at a 2-SD level is not one of the standard criteria. Assuming normal variation, approximately one in 20 QC results will be out of range, a high level of false positives. However, if run rejection requires two consecutive 2-SD errors, the base possibility of rejection is 0.25%, a lower percentage than that of the 3-SD cutoff. Relying on a QC to be abnormal and then repeat as abnormal requires a high level of systemic bias, and if the issue is increased variability, a repeated QC could very well be normal. Even if there is bias, the requirement for two consecutive QC measurements to be out at 2 SD may miss many low-lying biases that a 10x rule (which flags a run if 10 consecutive QC results are off by 1 SD in the same direction) would detect. In contrast, an increase in assay variability may be detected better by a 1-3S rule, which would flag a run where a single QC value is out by 3 SD. Of course, combining both of these rules would allow for detection of these two very different types of error. In this manner, multirule QC checks have the capacity to detect different errors at a higher sensitivity than a single blanket rule. Although not a large number of studies have validated the practice of repeating QC samples, a study in 2012 suggested that this process can provide performance on par with a 1-3S/2-2S/R-4S multirule, with the tradeoff of slightly increased cost of QC materials (due to increased rate of repetition).9 However, this study was limited by the use of a simple in silico model (that simulated differing levels of systemic error) and only compared with an abbreviated Westgard multirule. An additional level of variation that was not captured in this survey is the method of SD derivation. The most common method for calculating SD for QC purposes is to run a control analyte numerous times (usually at least 20 times) and measure the SD of the results. Depending on the laboratory, machine, and analyte, the number of repeats used to calculate the SD may vary. In addition, many controls have manufacturer-recommended standard deviation ranges as part of their package insert. The overwhelming majority of hospitals did not use moving averages (90%), although there was interest in implementing this in 19% of the hospitals. Only one hospital used moving averages as part of its QC practice (the other ran them in the background but did not fully use them), and one hospital had previously used moving averages but discontinued them due to perceived lack of utility. Overall, this is a surprising finding, as moving averages are theoretically useful, as well as inexpensive and simple to implement. The software to track these moving averages is included in commercially available software systems. Moving averages have the potential to detect low-level drifts in the values of measurements that would not ordinarily trigger normal QC flags.10-13 One laboratory mentioned that it used performance-driven QC methods, another proposed means of establishing QC goals.14 This involves using biological variation data to set goals for total allowable error and QC rules. For tests such as alanine aminotransferase, which have a high intraindividual and interindividual variation (24.3% and 41.6%, respectively),15 the goals for acceptable imprecision could theoretically be relaxed from ±2 SD to ±2.5 or even ±3 SD. This method may result in a reduction in the number of false rejections and maximize the number of true rejects. Overall, our findings demonstrate a heterogenous but surprisingly similar grouping of QC practices at these academic laboratories. At least 75% of the hospitals used a QC range of 2 SD, and virtually all (90%) used the policy of repeating an out-of-control QC and accepting results if the repeat value comes into control. The method of repeating out-of-control QC at 2 SD has been shown to be effective at improving the performance of QC over simple multirule methods (in this case, 1-3SD, 2 of 3-2SD/R-4) in an in silico model.9 In comparison to these QC methods developed through mathematical analysis, most of these policies appear to have evolved in a fashion “validated by experience.” The 2 SD or similar cutoff with QC repetition was seen commonly in our cohort, demonstrating that it is likely deeply ingrained in clinical pathology. Limitations of this study include the survey-based method: to ensure maximal compliance, we limited the number of items that we queried. Extremely detailed and granular information for each institution was not always available, so the data set used was not uniform, which creates the possibility of misinterpretation based on the answers. We attempted to minimize this by circulating the manuscript with the participating authors (each author was aware of which anonymized letter corresponded to his or her laboratory). Total quality management of a laboratory includes QC as a basic tenet but also a wide array of next-level practices to ensure test result quality. Future directions for this research include deeper-level inquiry into the heterogeneity of QC programs as well as investigating how these different practices affect result reporting and patient care. In conclusion, this study demonstrated both similarities and differences among QC practices at academic hospitals. There appears to be no systematic approach to defining QC rules or frequency. The Westgard rules offer a systematic and thoroughly vetted approach for QC to detect errors while minimizing false-positive rates, and additional methods involving rules for QC-level repetition have also been mathematically studied. Interestingly, Westgard rules were used by a small minority of academic center laboratories. Most laboratories prefer home-validated QC rules, which may rely on the director’s experience and expertise rather than a rigorously validated statistical approach to designing QC rules. We believe that most academic medical center chemistry laboratories that have similar volumes and patient populations may benefit from a standardized approach to QC for routine chemistry/immunochemistry testing and that can withstand rigorous statistical validation. The survey results suggest an opportunity for laboratory professional organizations to convene a consensus panel to determine a best practice approach (or approaches) to QC in the chemistry laboratory. This would help to ensure quality testing by enhancing error detection and reduce the costs associated with excessive QC and the use of QC rules that create high (false-positive) run rejection rates. References 1. Clinical Laboratory Improvement Amendments (CLIA). https://wwwn.cdc.gov/clia/Regulatory/default.aspx. Accessed January 1, 2017. 2. Kaplan AK,Pesce AJ. Clinical Chemistry: Theory, Analysis, Correlation . St Louis, MO: Mosby; 2009. 3. Burtis C, Ashwood E, Bruns D. Tietz Textbook of Clinical Chemistry and Molecular Diagnostics . Amsterdam, the Netherlands: Elsevier Saunders; 2006. 4. Mcpherson RA, Pincus MR. Henrys Clinical Diagnosis and Management by Laboratory Methods . Philadelphia, PA: Saunders; 2011. 5. Westgard JO, Barry PL, Hunt MRet al.   A multi-rule Shewhart chart for quality control in clinical chemistry. Clin Chem . 1981; 27: 493- 501. Google Scholar PubMed  6. Coskun A. Westgard multirule for calculated laboratory tests. Clin Chem Lab Med . 2006; 44: 1183- 1187. Google Scholar PubMed  7. Harder AC, Ben. 2016-17 Best hospitals honor roll and overview. 2016. http://health.usnews.com/health-care/best-hospitals/articles/best-hospitals-honor-roll-and-overview. Accessed January 1, 2017. 8. Westgard JO, Groth T, Aronsson Tet al.   Performance characteristics of rules for internal quality control: probabilities for false rejection and error detection. Clin Chem . 1977; 23: 1857- 1867. Google Scholar PubMed  9. Parvin CA, Kuchipudi L, Yundt-Pacheco JC. Should I repeat my 1:2s QC rejection? Clin Chem . 2012; 58: 925- 929. Google Scholar CrossRef Search ADS PubMed  10. Fleming JK, Katayev A. Changing the paradigm of laboratory quality control through implementation of real-time test results monitoring: for patients by patients. Clin Biochem . 2015; 48: 508- 513. Google Scholar CrossRef Search ADS PubMed  11. Wilson A, Roberts WL, Pavlov Iet al.   Patient result median monitoring for clinical laboratory quality control. Clin Chim Acta . 2011; 412: 1441- 1446. Google Scholar CrossRef Search ADS PubMed  12. Liu J, Tan CH, Badrick Tet al.   Moving sum of number of positive patient result as a quality control tool. Clin Chem Lab Med . 2017; 55: 1709- 1714. Google Scholar PubMed  13. Ng D, Polito FA, Cervinski MA. Optimization of a moving averages program using a simulated annealing algorithm: the goal is to monitor the process not the patients. Clin Chem . 2016; 62: 1361- 1371. Google Scholar CrossRef Search ADS PubMed  14. Brooks ZC. Performance-Driven Quality Control . Washington, DC: AACC Press; 2001. 15. Fraser CG. Biological Variation: From Principles to Practice. Washington, DC: AACC Press; 2001. © American Society for Clinical Pathology, 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

American Journal of Clinical PathologyOxford University Press

Published: May 29, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off