TY - JOUR
AU - King, Allison, R.
AB - Abstract Purpose The use of recommended practices for preventing and detecting diversion of prescription controlled substances at U.S. acute-care institutions, as reported by a sample of pharmacy service providers, were characterized. Methods A 41-item questionnaire was developed for an online survey of directors of pharmacy regarding strategies to combat controlled-substance diversion at their institutions. The survey questions were based on recommendations presented in a 2007 series of articles in the professional literature focusing on diversion control in three areas (the pharmacy, the operating room, and nursing units). Only institutions that had an accredited pharmacy residency program or were members of the University HealthSystem Consortium (UHC), an alliance of U.S. academic medical centers and affiliated hospitals, were targeted for the survey. Four hundred ninety-nine pharmacists were invited to participate in the survey, and 140 survey responses were received; all respondents did not answer all questions. Results The survey responses indicated considerable variation among the institutions in the use of 37 specific recommended practices, as reported by the pharmacy providers. Statistical analysis of comparative data suggested that larger institutions (400 or more licensed beds) were more likely to be using more of the recommended practices. Conclusion The results of a survey of directors of pharmacy at a sample of U.S. institutions (hospitals that had pharmacy residency programs or were UHC members) suggest wide variation in facilities’ use of recommended practices for the prevention and detection of controlled-substance diversion. Compliance, Controlled substances, Crime, Data collection, Guidelines, Pharmaceutical services, Pharmacy, institutional, hospital All health care institutions that handle controlled substances are required to develop storage and distribution systems to minimize the risk of diversion. The Drug Enforcement Administration (DEA) estimated that prescription drug diversion in the United States is a “$25 billion-a-year industry.”1 DEA reported that 563,677 methadone dosage units (10 mg/unit) and 226,519 OxyContin (oxycodone) tablets of all strengths were stolen during the period 2001–03, with 23.1% and 28.0% of those thefts, respectively, attributed to employee pilferage.2,3 Drug addiction among health care workers is well documented. In a survey conducted by Trinkoff et al.,4 nurses who reported a perception of easier availability were determined to be almost twice as likely as others to divert and use a controlled substance; nurses employed by institutions perceived to have poor workplace control of controlled substances were 1.5 times more likely to use a controlled substance. In another survey, 19% of pharmacists reported use of a controlled substance without a prescription during the preceding 12 months.5 In a study by Hughes et al.,6 17.6% of physicians surveyed reported the use of a minor opiate within the past year; of those, 92.5% cited self-treatment as the reason for use. Diversion is difficult to detect due to the numerous medication-access points embedded within most hospital distribution systems. Every transfer step throughout the dispensing process (e.g., transfer from point of purchase to pharmacy storage vault) should be assessed for the risk of diversion, and audit processes implemented at the most critical risk points. There are multiple strategies pharmacy departments can employ to prevent and detect diversion. Implementation of automated dispensing hardware and software is critically important to the diversion detection process. The widespread adoption of automated dispensing machines (ADMs) has greatly improved the physical security of controlled substances and enabled electronic documentation of dispensing, waste, and expiration processes. Software systems continue to evolve to meet the need for rapid and accurate identification of personnel with controlled-substance transaction counts that significantly exceed those of peers. Physical controls are also of great importance, as are operational policies that direct personnel to handle controlled substances in a manner that reduces the temptation and ease of diversion; examples include policies for handling of multidose vial waste and policies that ban personal belongings from all medication storage and fill areas. Currently there are no broadly accepted best practices or guidelines in the literature that institutions can adopt to improve controlled- substance diversion detection systems, but recommendations published in recent years in the professional literature can guide those efforts. A series of articles published in the journal Hospital Pharmacy focused on diversion of controlled substances in the acute-care setting.7–13 The series included numerous recommendations to reduce the risk of diversion or aid in its detection at key controlled- substance handling points in the pharmacy, procedure areas, and nursing units. In an effort to ascertain current health-system practices for detection of controlled-substance diversion, we compiled those recommendations into a questionnaire and contacted a national sample of directors of pharmacy to request their participation in an online survey. The primary objective of this research was to determine the prevalence of reported use of diversion detection practices recommended in the aforementioned series of articles. A secondary objective was to provide health-system pharmacists with a listing of these practices as an aid in assessment and improvement of their institutions’ diversion detection systems. Methods The survey questions were designed to identify variances from the recommended practices, as reported by a national sample of pharmacy directors. Initially, a pilot questionnaire was sent to pharmacists at three institutions to identify flaws and areas for improvement. The final survey consisted of 41 questions: 17 multiple-choice questions, 20 yes-or-no questions, and 4 questions about the recipients’ institution (size, ownership, location, and name). Through an online survey website, the survey link was e-mailed in April 2009 to directors of pharmacy identified through the pharmacy residency listings on the American Society of Health-System Pharmacists (ASHP) website (www.ashp.org) and the online member center of the University HealthSystem Consortium (UHC) (www.uhc.edu). UHC is an alliance whose membership includes about 90% of the nation’s nonprofit academic medical centers and about 250 affiliated hospitals. The initial e-mail described the research and provided the survey website link; two weeks later a reminder e-mail was sent. The survey remained open on the survey website for 30 days, and recipients were assured that their responses would be confidential. Directors of pharmacy at ASHP-accredited residency sites were selected with the aim of gathering data from a group representing current health-system pharmacy practice. Members of UHC were surveyed with the intent of including pharmacists who work for academic medical centers. The survey instructed recipients who were not directly involved in detection diversion efforts to forward the survey to another person in their organization whom they knew to have such involvement. The frequencies of reported use of surveyed practices were calculated by using descriptive statistics. Several questions allowed respondents to write comments; these were analyzed to determine if the responses were significant to the survey. Comparative data were analyzed using chi-square and Fisher’s exact tests. SPSS for Windows (SPSS Inc., Chicago, IL) was used for data analysis. The a priori level of significance was 0.05. Results The e-mail requesting participation in the survey was sent to 499 directors of pharmacy; 140 survey responses were received, an overall response rate of 28%. As 5 respondents did not provide information on the number of licensed beds, only 135 survey responses were included in the data analysis. Selected characteristics of survey respondents’ institutions are shown in Table 1. Of the respondents, 117 (87%) stated that their home institution was located in an urban area (population of >50,000); 95 (70%) reported private, not-for-profit ownership. Twenty-seven institutions (20%) were licensed for ≥700 beds, and 76 (56%) of the 135 institutions had at least 400 licensed beds. All survey respondents did not answer all survey questions; that is reflected in the data reporting (Table 2). Table 2 Compliance With Selected Recommended Practices for Diversion Detection and Preventiona Practice No. (%) Respondents Following Recommendation p Hospitals With 1–399 Beds (n = 59) Hospitals With ≥400 Beds (n = 76) Pharmacy  Witness and documentation required for wasting of controlled substances 57/59 (97) 72/76 (95) >0.05  If an expired-returns company is used, DEA form 222 provided by company is reconciled against central pharmacy vault inventory 48/54 (89) 62/71 (87) >0.05  Audit of controlled substance purchases against products added to pharmacy inventory performed at least quarterly 50/58 (86) 64/76 (84) >0.05  Automated vault for storage of controlled substances used 44/59 (75) 73/75 (97) 0.001  If CSOS not used, log of DEA form 222 kept to ensure all are accounted forb 34/43 (79) 52/58 (90) >0.05  Individual who submits a controlled-substance purchase order not responsible for receiving the order 43/59 (73) 66/76 (87) 0.049  Use and waste of multidose vials of controlled substances audited 39/57 (68) 58/74 (78) >0.05  Discrepancies reviewed by individual not routinely involved in controlled-substance handling 40/59 (68) 55/75 (73) >0.05  Person-to-person transfers of controlled substances audited, including transfers to nonautomated storage areas 48/59 (81) 46/75 (61) 0.014  Cameras directed at controlled-substance storage areas 32/58 (55) 54/74 (73) 0.043  Cameras directed at storage areas for high-cost noncontrolled medications 18/56 (32) 30/76 (39) >0.05  Purchases periodically reconciled against dispensings for high-cost noncontrolled medications 14/58 (24) 32/76 (42) 0.043  Personal belongings banned from drug storage areas 14/59 (24) 28/76 (37) >0.05  Packaging of high-cost noncontrolled medications defaced to inhibit diversion and resale 1/59 (2) 5/76 (7) . . .c Operating Rooms  Controlled substances dispensed for an OR case always reconciled by pharmacy against products documented as administered or returned to the pharmacy for wasting 26/59 (44) 48/76 (63) 0.036  Controlled substances not used in an OR case returned to the pharmacy for wasting 12/59 (20) 43/76 (57) 0.001 Nursing Units  Decentralized ADMs used for controlled-substance distribution 55/59 (93) 76/76 (100) . . .c  ADM discrepancy resolution explanations “always” or “sometimes” investigated for validity 54/55 (98) 70/76 (92) . . .c  Use of system capable of electronically identifying discrepancies between central pharmacy vault withdrawals and nursing unit-based ADM receipts 51/55 (93) 74/76 (97) . . .c  Use of “blind” count (ADM user forced to enter inventory count when accessing a pocket) 51/56 (91) 66/75 (88) >0.05  ADM discrepancies analyzed to identify individuals most frequently involved in discrepancies 38/57 (67) 55/76 (72) >0.05  ADM stock outages investigated for potential diversion 34/56 (61) 43/75 (57) >0.05  Biometric fingerprint scan used for ADM access 34/59 (58) 50/76 (66) >0.05  Locking cases used to secure non-PCA controlled-substance infusion containers while being administered 20/58 (34) 27/74 (36) >0.05  If diversion detection software used, flagged individuals are “always” or “sometimes” investigated for potential diversion 45/51 (88) 71/72 (99) . . .c  Diversion detection software flags users for controlled-substance ADM transaction counts significantly above peer group mean 40/58 (69) 66/76 (87) 0.018  Alert (e-mail, voice mail, pager) sent to pharmacy leaders when a specified ADM per-transaction threshold exceeded 14/53 (26) 25/71 (35) >0.05 Practice No. (%) Respondents Following Recommendation p Hospitals With 1–399 Beds (n = 59) Hospitals With ≥400 Beds (n = 76) Pharmacy  Witness and documentation required for wasting of controlled substances 57/59 (97) 72/76 (95) >0.05  If an expired-returns company is used, DEA form 222 provided by company is reconciled against central pharmacy vault inventory 48/54 (89) 62/71 (87) >0.05  Audit of controlled substance purchases against products added to pharmacy inventory performed at least quarterly 50/58 (86) 64/76 (84) >0.05  Automated vault for storage of controlled substances used 44/59 (75) 73/75 (97) 0.001  If CSOS not used, log of DEA form 222 kept to ensure all are accounted forb 34/43 (79) 52/58 (90) >0.05  Individual who submits a controlled-substance purchase order not responsible for receiving the order 43/59 (73) 66/76 (87) 0.049  Use and waste of multidose vials of controlled substances audited 39/57 (68) 58/74 (78) >0.05  Discrepancies reviewed by individual not routinely involved in controlled-substance handling 40/59 (68) 55/75 (73) >0.05  Person-to-person transfers of controlled substances audited, including transfers to nonautomated storage areas 48/59 (81) 46/75 (61) 0.014  Cameras directed at controlled-substance storage areas 32/58 (55) 54/74 (73) 0.043  Cameras directed at storage areas for high-cost noncontrolled medications 18/56 (32) 30/76 (39) >0.05  Purchases periodically reconciled against dispensings for high-cost noncontrolled medications 14/58 (24) 32/76 (42) 0.043  Personal belongings banned from drug storage areas 14/59 (24) 28/76 (37) >0.05  Packaging of high-cost noncontrolled medications defaced to inhibit diversion and resale 1/59 (2) 5/76 (7) . . .c Operating Rooms  Controlled substances dispensed for an OR case always reconciled by pharmacy against products documented as administered or returned to the pharmacy for wasting 26/59 (44) 48/76 (63) 0.036  Controlled substances not used in an OR case returned to the pharmacy for wasting 12/59 (20) 43/76 (57) 0.001 Nursing Units  Decentralized ADMs used for controlled-substance distribution 55/59 (93) 76/76 (100) . . .c  ADM discrepancy resolution explanations “always” or “sometimes” investigated for validity 54/55 (98) 70/76 (92) . . .c  Use of system capable of electronically identifying discrepancies between central pharmacy vault withdrawals and nursing unit-based ADM receipts 51/55 (93) 74/76 (97) . . .c  Use of “blind” count (ADM user forced to enter inventory count when accessing a pocket) 51/56 (91) 66/75 (88) >0.05  ADM discrepancies analyzed to identify individuals most frequently involved in discrepancies 38/57 (67) 55/76 (72) >0.05  ADM stock outages investigated for potential diversion 34/56 (61) 43/75 (57) >0.05  Biometric fingerprint scan used for ADM access 34/59 (58) 50/76 (66) >0.05  Locking cases used to secure non-PCA controlled-substance infusion containers while being administered 20/58 (34) 27/74 (36) >0.05  If diversion detection software used, flagged individuals are “always” or “sometimes” investigated for potential diversion 45/51 (88) 71/72 (99) . . .c  Diversion detection software flags users for controlled-substance ADM transaction counts significantly above peer group mean 40/58 (69) 66/76 (87) 0.018  Alert (e-mail, voice mail, pager) sent to pharmacy leaders when a specified ADM per-transaction threshold exceeded 14/53 (26) 25/71 (35) >0.05 a For fractional data, numerator denotes number reporting use; denominator denotes number answering survey item. DEA = Drug Enforcement Administration, CSOS = controlled-substances ordering system, OR = operating room, ADM = automated dispensing machine, PCA = patient-controlled analgesia. b Not applicable to 29 institutions that use electronic systems other than CSOS. c Data limitations precluded statistical analysis. Open in new tab Table 2 Compliance With Selected Recommended Practices for Diversion Detection and Preventiona Practice No. (%) Respondents Following Recommendation p Hospitals With 1–399 Beds (n = 59) Hospitals With ≥400 Beds (n = 76) Pharmacy  Witness and documentation required for wasting of controlled substances 57/59 (97) 72/76 (95) >0.05  If an expired-returns company is used, DEA form 222 provided by company is reconciled against central pharmacy vault inventory 48/54 (89) 62/71 (87) >0.05  Audit of controlled substance purchases against products added to pharmacy inventory performed at least quarterly 50/58 (86) 64/76 (84) >0.05  Automated vault for storage of controlled substances used 44/59 (75) 73/75 (97) 0.001  If CSOS not used, log of DEA form 222 kept to ensure all are accounted forb 34/43 (79) 52/58 (90) >0.05  Individual who submits a controlled-substance purchase order not responsible for receiving the order 43/59 (73) 66/76 (87) 0.049  Use and waste of multidose vials of controlled substances audited 39/57 (68) 58/74 (78) >0.05  Discrepancies reviewed by individual not routinely involved in controlled-substance handling 40/59 (68) 55/75 (73) >0.05  Person-to-person transfers of controlled substances audited, including transfers to nonautomated storage areas 48/59 (81) 46/75 (61) 0.014  Cameras directed at controlled-substance storage areas 32/58 (55) 54/74 (73) 0.043  Cameras directed at storage areas for high-cost noncontrolled medications 18/56 (32) 30/76 (39) >0.05  Purchases periodically reconciled against dispensings for high-cost noncontrolled medications 14/58 (24) 32/76 (42) 0.043  Personal belongings banned from drug storage areas 14/59 (24) 28/76 (37) >0.05  Packaging of high-cost noncontrolled medications defaced to inhibit diversion and resale 1/59 (2) 5/76 (7) . . .c Operating Rooms  Controlled substances dispensed for an OR case always reconciled by pharmacy against products documented as administered or returned to the pharmacy for wasting 26/59 (44) 48/76 (63) 0.036  Controlled substances not used in an OR case returned to the pharmacy for wasting 12/59 (20) 43/76 (57) 0.001 Nursing Units  Decentralized ADMs used for controlled-substance distribution 55/59 (93) 76/76 (100) . . .c  ADM discrepancy resolution explanations “always” or “sometimes” investigated for validity 54/55 (98) 70/76 (92) . . .c  Use of system capable of electronically identifying discrepancies between central pharmacy vault withdrawals and nursing unit-based ADM receipts 51/55 (93) 74/76 (97) . . .c  Use of “blind” count (ADM user forced to enter inventory count when accessing a pocket) 51/56 (91) 66/75 (88) >0.05  ADM discrepancies analyzed to identify individuals most frequently involved in discrepancies 38/57 (67) 55/76 (72) >0.05  ADM stock outages investigated for potential diversion 34/56 (61) 43/75 (57) >0.05  Biometric fingerprint scan used for ADM access 34/59 (58) 50/76 (66) >0.05  Locking cases used to secure non-PCA controlled-substance infusion containers while being administered 20/58 (34) 27/74 (36) >0.05  If diversion detection software used, flagged individuals are “always” or “sometimes” investigated for potential diversion 45/51 (88) 71/72 (99) . . .c  Diversion detection software flags users for controlled-substance ADM transaction counts significantly above peer group mean 40/58 (69) 66/76 (87) 0.018  Alert (e-mail, voice mail, pager) sent to pharmacy leaders when a specified ADM per-transaction threshold exceeded 14/53 (26) 25/71 (35) >0.05 Practice No. (%) Respondents Following Recommendation p Hospitals With 1–399 Beds (n = 59) Hospitals With ≥400 Beds (n = 76) Pharmacy  Witness and documentation required for wasting of controlled substances 57/59 (97) 72/76 (95) >0.05  If an expired-returns company is used, DEA form 222 provided by company is reconciled against central pharmacy vault inventory 48/54 (89) 62/71 (87) >0.05  Audit of controlled substance purchases against products added to pharmacy inventory performed at least quarterly 50/58 (86) 64/76 (84) >0.05  Automated vault for storage of controlled substances used 44/59 (75) 73/75 (97) 0.001  If CSOS not used, log of DEA form 222 kept to ensure all are accounted forb 34/43 (79) 52/58 (90) >0.05  Individual who submits a controlled-substance purchase order not responsible for receiving the order 43/59 (73) 66/76 (87) 0.049  Use and waste of multidose vials of controlled substances audited 39/57 (68) 58/74 (78) >0.05  Discrepancies reviewed by individual not routinely involved in controlled-substance handling 40/59 (68) 55/75 (73) >0.05  Person-to-person transfers of controlled substances audited, including transfers to nonautomated storage areas 48/59 (81) 46/75 (61) 0.014  Cameras directed at controlled-substance storage areas 32/58 (55) 54/74 (73) 0.043  Cameras directed at storage areas for high-cost noncontrolled medications 18/56 (32) 30/76 (39) >0.05  Purchases periodically reconciled against dispensings for high-cost noncontrolled medications 14/58 (24) 32/76 (42) 0.043  Personal belongings banned from drug storage areas 14/59 (24) 28/76 (37) >0.05  Packaging of high-cost noncontrolled medications defaced to inhibit diversion and resale 1/59 (2) 5/76 (7) . . .c Operating Rooms  Controlled substances dispensed for an OR case always reconciled by pharmacy against products documented as administered or returned to the pharmacy for wasting 26/59 (44) 48/76 (63) 0.036  Controlled substances not used in an OR case returned to the pharmacy for wasting 12/59 (20) 43/76 (57) 0.001 Nursing Units  Decentralized ADMs used for controlled-substance distribution 55/59 (93) 76/76 (100) . . .c  ADM discrepancy resolution explanations “always” or “sometimes” investigated for validity 54/55 (98) 70/76 (92) . . .c  Use of system capable of electronically identifying discrepancies between central pharmacy vault withdrawals and nursing unit-based ADM receipts 51/55 (93) 74/76 (97) . . .c  Use of “blind” count (ADM user forced to enter inventory count when accessing a pocket) 51/56 (91) 66/75 (88) >0.05  ADM discrepancies analyzed to identify individuals most frequently involved in discrepancies 38/57 (67) 55/76 (72) >0.05  ADM stock outages investigated for potential diversion 34/56 (61) 43/75 (57) >0.05  Biometric fingerprint scan used for ADM access 34/59 (58) 50/76 (66) >0.05  Locking cases used to secure non-PCA controlled-substance infusion containers while being administered 20/58 (34) 27/74 (36) >0.05  If diversion detection software used, flagged individuals are “always” or “sometimes” investigated for potential diversion 45/51 (88) 71/72 (99) . . .c  Diversion detection software flags users for controlled-substance ADM transaction counts significantly above peer group mean 40/58 (69) 66/76 (87) 0.018  Alert (e-mail, voice mail, pager) sent to pharmacy leaders when a specified ADM per-transaction threshold exceeded 14/53 (26) 25/71 (35) >0.05 a For fractional data, numerator denotes number reporting use; denominator denotes number answering survey item. DEA = Drug Enforcement Administration, CSOS = controlled-substances ordering system, OR = operating room, ADM = automated dispensing machine, PCA = patient-controlled analgesia. b Not applicable to 29 institutions that use electronic systems other than CSOS. c Data limitations precluded statistical analysis. Open in new tab Table 1 Demographic Characteristics of Survey Respondents (n = 135) Characteristic No. (%) Respondents Hospitals With 1–399 Beds (n = 59) Hospitals With ≥400 Beds (n = 76) Hospital settinga  Rural (population ≤50,000) 16 (27) 2 (3)  Urban (population >50,000) 43 (73) 74 (97) Type of institutiona  Private, for-profit 1 (2) 5 (7)  Private, not-for-profit 34 (58) 61 (80)  State or government 24 (41) 10 (13) Characteristic No. (%) Respondents Hospitals With 1–399 Beds (n = 59) Hospitals With ≥400 Beds (n = 76) Hospital settinga  Rural (population ≤50,000) 16 (27) 2 (3)  Urban (population >50,000) 43 (73) 74 (97) Type of institutiona  Private, for-profit 1 (2) 5 (7)  Private, not-for-profit 34 (58) 61 (80)  State or government 24 (41) 10 (13) a Significant difference between small and large hospitals (p < 0.05). Open in new tab Table 1 Demographic Characteristics of Survey Respondents (n = 135) Characteristic No. (%) Respondents Hospitals With 1–399 Beds (n = 59) Hospitals With ≥400 Beds (n = 76) Hospital settinga  Rural (population ≤50,000) 16 (27) 2 (3)  Urban (population >50,000) 43 (73) 74 (97) Type of institutiona  Private, for-profit 1 (2) 5 (7)  Private, not-for-profit 34 (58) 61 (80)  State or government 24 (41) 10 (13) Characteristic No. (%) Respondents Hospitals With 1–399 Beds (n = 59) Hospitals With ≥400 Beds (n = 76) Hospital settinga  Rural (population ≤50,000) 16 (27) 2 (3)  Urban (population >50,000) 43 (73) 74 (97) Type of institutiona  Private, for-profit 1 (2) 5 (7)  Private, not-for-profit 34 (58) 61 (80)  State or government 24 (41) 10 (13) a Significant difference between small and large hospitals (p < 0.05). Open in new tab Ninety-seven percent of respondents (131 of 135) indicated that their institution was using decentralized ADMs, but almost 13% of institutions (17 of 134 respondents) reported the lack of an automated controlled-substance storage vault within the central pharmacy. Sixty-five percent of respondents (86 of 132) indicated the presence of cameras directed at controlled-substance storage areas in the pharmacy (only 36% of respondents [48 of 132] reported the use of cameras directed at storage locations for high-cost non-controlled medications). Only 31% of institutions (42 of 135) reported an existing policy banning all personal belongings (e.g., purses, backpacks) from drug storage areas. Four percent of respondents (6 of 135) indicated that they defaced boxes of high-cost medications (e.g., removed box top, discolored packaging with a marker) to inhibit their diversion and subsequent resale. The most commonly reported method for purchasing controlled substances included having one person create and submit a controlled-substance purchase order and another person receive the purchase (109 of 135, or 81% of respondents). Approximately half of the respondents (71 of 133) reported that they allowed the purchaser to have access to controlled substances beyond what is necessary to create reports. Twenty-nine respondents (22%) were using the DEA controlled- substance ordering system, which allows for electronic ordering of schedule II substances. Excluding these institutions, 85% of respondents (86 of 101) maintained a log of DEA 222 forms to ensure that all forms are accounted for. Approximately 85% of respondents (114 of 134) reported performing an audit of controlled-substance purchases against inventory records at least quarterly. Seventy-four percent of respondents (97 of 131) reported auditing the use of and waste procedures for controlled-substance multidose vials to ensure accountability for all product; 70% (94 of 134) reported auditing person-to-person transactions (e.g., dispensing of a single fentanyl patch directly to a nurse) and other transfers to non-automated storage areas. Among institutions that employed an expired- returns company, reconciliation of records against returns to the company occurred in 88% of institutions. Only 34% (46 of 134) audit purchases of high-cost, noncontrolled medications against dispensation. Operating room practices Fifty-one percent of all respondents (69 of 135) reported use of a satellite pharmacy for dispensing to operating room (OR) suites. Of those, 12% (8 of 69) reported that the satellite pharmacy was staffed at all hours. Fifty-five percent of respondents (74 of 135) reported that their facilities always reconciled medication quantities documented on the dispensing record against the quantities documented as administered or returned to the pharmacy as waste, while 36% (49 of 135) indicated such audits were performed sometimes. Twenty-four percent (32 of 135) of respondents reported that unused controlled substances were not returned from the OR to the pharmacy. Among respondents receiving returns of unused controlled substances, 46 said their institutions did not test to verify the drugs’ identity. Nursing unit practices Ninety-five percent of respondents (125 of 131) reported that their facilities had a system capable of electronically identifying discrepancies between withdrawals from a controlled- substance vault in the central pharmacy and controlled substances received by an ADM on the nursing unit. For institutions using decentralized ADMs, 89% performed a “blind” count (the user was forced to enter the inventory count when accessing a controlled-substance pocket). Controlled-substance counts or inventories at an ADM were performed at each transaction by 30% of respondents (41 of 135), at each nursing shift change by 19% (26 of 135), daily by 9% (12 of 135), weekly by 33% (44 of 135), and at some other frequency by 9% (12 of 135). About 62% of respondents (84 of 135) reported use of biometric fingerprint-scanning technology for ADM access. About 69% of respondents (93 of 135) reported that trends in controlled-substance audit discrepancies were monitored for investigation of users most frequently involved in discrepancy creation; flagged discrepancies often required the user to provide a reason at the point of resolution, and 92% of respondents (124 of 131) indicated that explanations were “always” or “sometimes” investigated for validity. About 59% of respondents (77 of 131) said ADM stock outages were investigated for potential diversion at least monthly. Seventy-nine percent of respondents (106 of 134) reported use of diversion detection software capable of flagging users whose controlled- substance ADM withdrawals exceeded a defined threshold (e.g., 3 standard deviations above the mean number of controlled-substance transactions among peers). Software used by survey respondents is shown in Table 3. Table 3 Diversion Detection Software Software Source No. (%) Respondents Hospitals With 1–399 Beds (n = 41) Hospitals With ≥400 Beds (n = 65) Vendor of automated distribution software 25 (61) 31 (48) Pandora Data Systems (Scotts Valley, CA) 8 (20) 15 (23) Rx Auditor (Medacist Solutions Group, Cheshire, CT) 4 (10) 12 (19) System developed in institution 2 (5) 3 (5) Other 2 (5) 4 (6) Software Source No. (%) Respondents Hospitals With 1–399 Beds (n = 41) Hospitals With ≥400 Beds (n = 65) Vendor of automated distribution software 25 (61) 31 (48) Pandora Data Systems (Scotts Valley, CA) 8 (20) 15 (23) Rx Auditor (Medacist Solutions Group, Cheshire, CT) 4 (10) 12 (19) System developed in institution 2 (5) 3 (5) Other 2 (5) 4 (6) Open in new tab Table 3 Diversion Detection Software Software Source No. (%) Respondents Hospitals With 1–399 Beds (n = 41) Hospitals With ≥400 Beds (n = 65) Vendor of automated distribution software 25 (61) 31 (48) Pandora Data Systems (Scotts Valley, CA) 8 (20) 15 (23) Rx Auditor (Medacist Solutions Group, Cheshire, CT) 4 (10) 12 (19) System developed in institution 2 (5) 3 (5) Other 2 (5) 4 (6) Software Source No. (%) Respondents Hospitals With 1–399 Beds (n = 41) Hospitals With ≥400 Beds (n = 65) Vendor of automated distribution software 25 (61) 31 (48) Pandora Data Systems (Scotts Valley, CA) 8 (20) 15 (23) Rx Auditor (Medacist Solutions Group, Cheshire, CT) 4 (10) 12 (19) System developed in institution 2 (5) 3 (5) Other 2 (5) 4 (6) Open in new tab Discussion Preventing diversion of controlled substances is a challenging process for hospitals nationwide. The survey findings suggest varying levels of use of recommended diversion prevention practices at institutions represented in the survey. Currently, there are few reports or guidelines on the use of strategies to prevent diversion in the professional literature. Although 94% of respondents reported use of ADMs, 16% did not use an automated controlled substance vault. This raises the question of how these institutions securely stored controlled substances in the pharmacy. Automated controlled substance vaults allow users to detect diversion through reporting software, maintain a perpetual inventory, and provide electronic documentation of all transactions. More than one third of institutions did not use cameras directed at controlled- substance storage areas. Cameras offer a relatively low cost:value ratio for detecting diversion, and even nonfunctioning cameras can offer some deterrent value. The survey results suggest that the greatest opportunities for preventing diversion lie in the purchasing and storage of controlled substances. Most respondent institutions had one person purchase and another person receive controlled substances, which allows for tighter control of inventory through a system of checks and balances. About half of respondents did not allow the purchaser access to controlled substances beyond the access necessary to prepare reports for ordering purposes, but such restricted access is not always feasible. Smaller departments, for example, may have only one staff member to purchase and receive controlled substances and stock the drugs in ADMs on nursing units. Most institutions did not ban all personal belongings from drug storage areas, so there is a possibility that someone could place items in a purse or backpack to be taken out of the pharmacy at the end of a shift. Auditing of controlled substance purchases helps ensure alignment of purchases with actual inventory. Most respondents reported maintaining a DEA form 222 log, but 12% of institutions did not conduct audits and thus might be at higher risk for diversion. Person-to-person transfers of controlled substances were audited by 69% of institutions, but such transfers can be difficult to monitor. One fourth of the respondents reported that they did not audit the use and waste of multidose vials of controlled substances; those institutions would be well advised to begin such auditing. There is a high risk of controlled-substance diversion in the OR. We found that 51% of institutions had an OR pharmacy, and 55% of institutions reported always having a pharmacy staff member reconcile the medication dispensed against those documented as administered or returned to the pharmacy for wasting. About 9% of respondent institutions reported that pharmacy staff members were not involved in the reconciliation of OR controlled substances. Although there are barriers to this process (e.g., small pharmacy staff, small OR, budget constraints), pharmacy departments should strive to procure the necessary resources for this important function. The pharmacy is accountable for all medications, whether stored in the pharmacy itself or in an ADM on a nursing unit. In 61% of respondent institutions, biometric fingerprint technology was used as the positive-identification system to control access to medications from an ADM. Biometric technology is the recommended identification system—nothing is more individualized than a person’s fingerprint—and its use should be increased. Not all of the institutions reporting use of ADMs also reported use of ADM software to identify controlled substance diversion. There may be inherent deficiencies in the software, but some investigation is better than none. Available software can identify users who are frequently involved in discrepancies involving controlled and non-controlled substances and can limit the quantity of controlled substances to which individual users are allowed access. Of the institutions that reported using basic software packages, 88% had purchased add-on software for more comprehensive diversion detection. ADM pockets within the drawers can only hold a certain quantity of medications. A stock-out notification is typically sent to the pharmacy via the ADM when the quantity reaches a critical level, before the pocket is entirely depleted. The survey results indicated that stock-out notifications were immediately investigated by 42% of institutions for potential diversion, but 34% reported never conducting such investigations. Institutions should take an active role in investigating stock-outs as an important method for prevention or detection of diversion. One major limitation to our study was a low response rate (28%). We suspect that many survey recipients ignored the e-mailed request to participate or forwarded it to another individual with more intimate knowledge of the questions asked. The survey’s length (41 questions) and the limited response period (30 days) also might have contributed to the low response rate. Despite those limitations, we believe that the distribution of institution size suggests a good sample mix. While this study provides information on diversion prevention, the study was limited to hospitals that were members of UHC or had residency programs listed on the ASHP website. More research needs to be conducted to better define current practices and develop national standards of practice for prevention of controlled-substance diversion. Conclusion The results of a survey of directors of pharmacy at a sample of U.S. institutions (hospitals that had pharmacy residency programs or were UHC members) suggest wide variation in facilities’ use of recommended practices for the prevention and detection of controlled-substance diversion. Footnotes The authors have declared no potential conflicts of interest. References 1 Inciardi J Surratt H Kurtz S et al. . Mechanisms of prescription drug diversion among drug-involved club- and street-based populations . Pain Med. 2007 ; 8 : 171 – 83 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Drug Enforcement Administration . Drug theft & loss . www.aatod.org/pdfs/methadone-theft.pdf (accessed 2010 Dec 21). 3 Drug Enforcement Administration . Oxycontin theft & loss incidents: January 2000 through June 2003 . www.deadiversion.usdoj.gov/drugs_concern (accessed 2010 Dec 21). 4 Trinkoff AM Storr CL Wall MP . Prescription-type drug misuse and work-place access among nurses . J Addict Dis. 1999 ; 18 (1) : 9 – 17 . Google Scholar Crossref Search ADS PubMed WorldCat 5 McAuliffe W Santangelo S Gingras J et al. . Use and abuse of controlled substances by pharmacists and pharmacy students . Am J Hosp Pharm. 1987 ; 44 : 311 – 7 . Google Scholar PubMed WorldCat 6 Hughes PH Brandenburg N Baldwin DC et al. . Prevalence of substance use among US physicians . JAMA. 1992 ; 267 : 2333 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 7 O’Neal B . Controlled substance diversion detection: go the extra mile . Hosp Pharm. 2004 ; 39 : 868 – 70 . Google Scholar Crossref Search ADS WorldCat 8 Siegel J O’Neal B . Diversion in the pharmacy . Hosp Pharm. 2007 ; 42 : 145 – 8 . Google Scholar Crossref Search ADS WorldCat 9 Siegel J O’Neal B . Code N: multidisciplinary approach to proactive drug diversion prevention . Hosp Pharm. 2007 ; 42 : 244 – 8 . Google Scholar Crossref Search ADS WorldCat 10 O’Neal B Bass K Siegel J . Diversion in the operating room . Hosp Pharm. 2007 ; 42 : 364 – 7 . WorldCat 11 Siegel J Wierwille C O’Neal B . The investigative process . Hosp Pharm. 2007 ; 42 : 466 – 9 . Google Scholar Crossref Search ADS WorldCat 12 O’Neal B Siegel J . Use of diversion detection software . Hosp Pharm. 2007 ; 42 : 564 – 71 . Google Scholar Crossref Search ADS WorldCat 13 Siegel J O’Neal B . Code N: the intervention process . Hosp Pharm. 2007 ; 42 : 653 – 6 . Google Scholar Crossref Search ADS WorldCat Copyright © 2011, American Society of Health-System Pharmacists, Inc. All rights reserved.
TI - Compliance with recommendations for prevention and detection of controlled-substance diversion in hospitals
JF - American Journal of Health-System Pharmacy
DO - 10.2146/ajhp100212
DA - 2011-04-15
UR - https://www.deepdyve.com/lp/oxford-university-press/compliance-with-recommendations-for-prevention-and-detection-of-840vbz2P0b
SP - 689
VL - 68
IS - 8
DP - DeepDyve
ER -