Predictors of latent tuberculosis infection treatment completion in the US private sector: an analysis of administrative claims data

Predictors of latent tuberculosis infection treatment completion in the US private sector: an... Background: Factors that affect latent tuberculosis infection (LTBI) treatment completion in the US have not been well studied beyond public health settings. This gap was highlighted by recent health insurance-related regulatory changes that are likely to increase LTBI treatment by private sector healthcare providers. We analyzed LTBI treatment completion in the private healthcare setting to facilitate planning around this important opportunity for tuberculosis (TB) control in the US. Methods: We analyzed a national sample of commercial insurance medical and pharmacy claims data for people ages 0 to 64 years who initiated daily dose isoniazid treatment between July 2011 and March 2014 and who had complete data. All individuals resided in the US. Factors associated with treatment completion were examined using multivariable generalized ordered logit models and bivariate Kruskal-Wallis tests or Spearman correlations. Results: We identified 1072 individuals with complete data who initiated isoniazid LTBI treatment. Treatment completion was significantly associated with less restrictive health insurance, age < 15 years, patient location, use of interferon-gamma release assays, non-poverty, HIV diagnosis, immunosuppressive drug therapy, and higher cumulative counts of clinical risk factors. Conclusions: Private sector healthcare claims data provide insights into LTBI treatment completion patterns and patient/provider behaviors. Such information is critical to understanding the opportunities and limitations of private healthcare in the US to support treatment completion as this sector’s role in protecting against and eliminating TB grows. Keywords: Latent tuberculosis infection, LTBI, Treatment completion, Claims data, Administrative data, Isoniazid, Epidemiology, Health service delivery, Public health practice, Medication adherence Background develop TB over time, with higher progression risk among Up to 13 million people in the US have latent tuberculosis immunocompromised persons [3]. Although LTBI treat- infection (LTBI) [1, 2]. These people are infected with ment does not eliminate the risk of progression to active Mycobacterium tuberculosis yet do not have active tubercu- TB, completion of a proven LTBI treatment regimen (e.g., losis (TB) disease; they are asymptomatic and cannot trans- 6 or 9 months of daily isoniazid, 4 months of daily rifam- mitTB.Withouttreatment5–10% of people with LTBI will pin, 12 doses of weekly isoniazid and rifapentine) dramat- ically decreases that risk [4]. The US’ strategic plan to eliminate domestic TB includes risk-targeted identification * Correspondence: Erica.Stockbridge@unthsc.edu Department of Health Behavior and Health Systems, University of North and treatment of people with LTBI [5]. This strategy is Texas Health Science Center School of Public Health, 3500 Camp Bowie Blvd, supported by the US Preventive Services Task Force’s Fort Worth, TX 76107, USA (USPSTF) recent “Grade B” rating for LTBI testing in Department of Advanced Health Analytics and Solutions, Magellan Health, Inc., 4800 N Scottsdale Rd #4400, Scottsdale, AZ 85251, USA high-risk populations, which indicates to primary care Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Stockbridge et al. BMC Public Health (2018) 18:662 Page 2 of 13 providers that targeted LTBI testing and treatment afford indicate whether the 6 or 9-month regimen was prescribed, moderate health benefit with little risk [6, 7]. we could determine how many doses of isoniazid were dis- Public health agencies have traditionally provided most pensed. Thus, we grouped isoniazid treatments into three TB control and prevention services in the US [8–11]. How- mutually exclusive ordinal categories: 1) non-completion ever, the USPSTF’s rating and current policy will likely drive (< 180 doses received within 9-months), 2) completion of increased involvement by private sector providers as health the 6-month regimen but not the 9-month regimen (180 to insurers are now required to cover TB/LTBI testing in 269 doses received within 9-months), or 3) completion of high-risk populations with no patient cost sharing [12]. At the 9-month regimen (≥ 270 doses received within the same time, the uninsured rate in the US is decreasing 12-months) [20]. These increasing levels of completion are [13] and health insurance coverage is associated with in- important because, while isoniazid treatment completion at creased use of primary and other private sector health care any duration does not necessarily imply LTBI cure, the risk [14]. These shifts present an opportunity to coordinate of progression to active TB decreases as the duration of iso- public/private approaches to TB prevention. Factors associ- niazid treatment increases [22]. ated with LTBI treatment completion are seldom studied outside of public health settings [15, 16]. Differences in pa- Explanatory variables tient risks, provider and patient incentives, and care pro- Explanatory variables were constructed from the medical cesses in the private sector suggest a need for more and pharmacy claims data (see Additional file 1 for details). information about the factors associated with treatment Socio-demographic variables included sex, age, census re- completion in this increasingly important arena. gion, and a patient location variable based on the National We used a national sample of commercial claims data Center for Health Statistics urban-rural classification [23]. to examine private sector LTBI treatment across the US The percentage of households living under the federal pov- as a step toward filling this gap. Insurance claims offer a erty level in a patient’s county served as a proxy for house- window into private healthcare practice patterns [17]. hold income [24]. Additional variables included insurance We aimed to use these data to identify factors associated type (health maintenance organization [HMO], point of with the completion of daily dose isoniazid LTBI treat- service [POS], or preferred provider organization [PPO]), ment in the private sector setting. prescription size (the supply of isoniazid received when the first prescription was filled; < 2 months or ≥2months), Methods year, and the type of LTBI diagnostic test received in the Data source and analytic sample 6 months before treatment initiation. Non-clinical variables We analyzed de-identified medical and pharmacy claims related to risk of LTBI or progression to active TB were in- from Optum Clinformatics® Data Mart (formerly called cluded, such as the state TB rate. While country of birth the National Research Database) which includes claims was unavailable, we included prevalence of foreign-born in- for approximately 30.6 million commercially insured indi- dividuals in the patient’s county as a proxy [25, 26]. Clinical viduals – about 19% of the commercially insured US risk factors included diabetes, tobacco use, HIV, immuno- population [18]. We analyzed data for a randomly selected suppressive medication use, contact with or exposure to sample of 4 million people who were ages 0 to 64 years. TB, and a history of or late effects of TB [27]. A simple Additional details about this sample are described else- count of each patient’s clinical risk factors represented cu- where [19]. We used a claims-based algorithm to identify mulative risk (i.e., 0, 1, or > 1 risk factor). individual 6 to 9 month daily dose isoniazid courses of treatment for LTBI [19], which have been the most com- Statistical analyses monly used LTBI treatment regimens [20]. We examined We calculated the proportion of individuals in each of treatment initiated between July 2011 and March 2014. In three categories of treatment completion (i.e., < 6 months, addition to requiring that data be available to determine if 6 to < 9 months, ≥9 months) and examined the bivariate treatment was completed (as specified in the algorithm) relationships between the explanatory variables and com- [19] we required non-missing socio-demographic variables pletion using Kruskal-Wallis tests and Spearman correla- (i.e., the percent of foreign-born in county, patient loca- tions. We explored the adjusted association between these tion category, percent of households in county living variables and treatment completion category using multi- under the federal poverty level (FPL), and state TB rate). variable generalized ordered logit models. Variables meet- ing the parallel-lines assumption were constrained to have Measures equal effects; the odds ratios for non-completion versus Outcome variable completing ≥6 months of treatment and those for com- The outcome of interest was completion of daily isoniazid pleting < 9 months of treatment versus ≥9 months of treatment for LTBI [21]. Patients maybeprescribeda 6or treatment were the same. Variables violating the assump- 9-month isoniazid regimen [4]. While our data do not tion were not constrained and consequently have different Stockbridge et al. BMC Public Health (2018) 18:662 Page 3 of 13 odds ratios for completion category comparisons [28]. We Table 1 Completion of daily-dose isoniazid treatment for latent tuberculosis infection. N = 1072 ran two multivariable generalized ordered logit models. In Model 1 we examined the relationship between completion Isoniazid Treatment Completion Number % of 95% Confidence Total Interval and cumulative risk. Model 2 explored the relationship be- Less than 6 months (Incomplete 577 53.82 50.82–56.79 tween completion and individual clinical risk factors. treatment) We also ran a multivariable logit model with completion At least 6 months 495 46.18 43.20–49.17 of ≥6 months of treatment as the outcome measure and ≥6 months but < 9 months 253 23.60 21.15–26.24 all predictors from the more detailed Model 2 as explana- tory variables. This logit model was used to examine the ≥9 months 242 22.57 20.17–25.18 reduction of variance in the treatment completion variable attributable to each predictor, which provided insight into the importance of the variables with respect to model pre- Tables 2 and 3 describe relationships between the ex- dictions of completing ≥6months oftreatment [29, 30]. planatory variables and the likelihood of treatment com- We conducted two sets of post hoc analyses. First, in pletion from bivariate analyses and multivariable models, order to assess the robustness of our findings we conducted respectively. Significant unadjusted non-clinical factors as- sensitivity analyses using variations on our treatment com- sociated with completion included younger age, PPO in- pletion outcomes measure. We ran four multivariable logis- surance, larger prescription size, and residing in a county tic regression models to explore characteristics associated with < 15% of households below FPL. Similarly, in the with completion of ≥5 months of treatment and compare multivariable models younger people (ages 0 to 14 years) the results to the characteristics associated with ≥6months had higher adjusted odds of treatment completion than of treatment in Models 1 and 2. Four models were used be- older people. Compared to people in large central metro- cause we had two sets of explanatory variables (see descrip- politan counties, those in large fringe metropolitan coun- tions of Models 1 and 2 above), and we defined completion ties had lower adjusted odds of completing ≥6monthsof two ways: 1) 150 doses in 9 months, and 2) 150 doses in treatment, although this association was not seen with 8 months. We explored the data using two definitions be- completing ≥9 months of treatment. Residing in a county cause we identified no previous studies or clinical practice with ≥15% of households below FPL was significantly as- guidelines defining a time period in which 150 doses sociated with lower adjusted odds of completion. Detailed (5 months) of isoniazid would be considered completed adjusted odds ratios for the associations described above treatment. are found in Table 3. Second, we explored our findings related to the LTBI Insurance type and prescription size were also signifi- testing variable. We ran a frequency distribution which cantly associated with completion. The adjusted odds of a contained additional details about the LTBI tests re- PPO-insured patient completing ≥6 months of treatment ceived. Additionally, to clarify differences between the were 1.8 to 1.9 times that of an HMO-insured patient, and results in our bivariate and multivariable analyses, we the odds of a PPO-insured patient completing ≥9months conducted post hoc bivariate analyses exploring the rela- were 2.8 to 2.9 times that of an HMO-insured patient. Lar- tionship between the explanatory variables and the type ger prescription size was associated with higher adjusted of LTBI diagnostic test using chi square tests for cat- odds of completing ≥9 months of treatment, although this egorical variables and ANOVAs for continuous variables. association was not seen for completing ≥6monthsof We used Stata 14.2 for most statistical testing [31]but treatment. used IBM SPSS Modeler 17 to complete the importance IGRA testing, HIV, and immunosuppressive medica- analysis [32]. All statistical testing was two-sided, and sig- tion use each had statistically significant bivariate associ- nificance was tested at p <.05. ations with treatment completion. In the multivariable model, people with HIV had an adjusted 2.5 times greater odds of an increased level of completion relative Results to those without. Additionally, both unadjusted and ad- Two (0.2%) of 1074 individuals identified with the justed likelihood of completion was significantly associ- algorithm as having initiated isoniazid LTBI treat- ated with cumulative clinical risk. Compared to people ment were excluded due to missing geographic vari- with no clinical risk factors, those with one risk factor ables. Of the remaining 1072 almost half (46.2%) had 1.5 times greater adjusted odds and those with more completed ≥6 months of treatment. The balance than one risk factor had 1.8 times greater adjusted odds (53.8%) initiated but did not complete the minimum of an increased level of treatment completion. The im- 6-months course. Roughly equal proportions com- portance analysis indicated that the most important vari- pleted ≥6 but < 9 months treatment or ≥9months able in predicting treatment of ≥6 months of treatment was (23.6 and 22.6% of all patients, respectively; Table 1). patient location, followed closelybyimmunosuppressive Stockbridge et al. BMC Public Health (2018) 18:662 Page 4 of 13 Table 2 Frequency distribution of patient characteristic variables for people initiating daily-dose isoniazid treatment and the proportion of people completing treatment by each characteristic. Treatment completion was categorized as 1) less than 6 months completed, 2) at least 6 months but less than 9 months completed, and 3) 9 or more months completed Distribution % Achieving Each Level of Isoniazid Treatment % Completing ≥6 Mo. Completion N % or Mean < 6 Months ≥6 but < ≥9 Months p-value: 3 ≥6 Months p-value: < 6 of Total Complete 9 Months Complete Completion Complete vs [% or Mean] Complete [% or Levels [% or ≥6 Months [% or Mean] Mean] Complete Mean] Sex Female 575 53.6% 55.8% 22.1% 22.1% 0.232 44.2% 0.158 Male 497 46.4% 51.5% 25.4% 23.1% 48.5% Age Group 0–14 105 9.8% 43.8% 24.8% 31.4% 0.019 56.2% 0.064 15–29 291 27.1% 58.8% 23.4% 17.9% 41.2% 30–44 321 29.9% 53.9% 25.2% 20.9% 46.1% 45–64 355 33.1% 52.7% 22.0% 25.3% 47.3% Census Region Northeast 352 32.8% 54.8% 20.5% 24.7% 0.148 45.2% 0.151 Midwest 174 16.2% 52.3% 25.3% 22.4% 47.7% South 148 13.8% 61.5% 22.3% 16.2% 38.5% West 398 37.1% 53.8% 23.6% 22.6% 46.2% Patient Location Large 484 45.1% 50.0% 26.7% 23.4% 0.169 50.0% 0.066 central metro county Large fringe 413 38.5% 57.6% 19.6% 22.8% 42.4% metro county Any smaller 175 16.3% 55.4% 24.6% 20.0% 44.6% county % of Households Under < 15% 596 55.6% 51.7% 22.8% 25.5% 0.035 48.3% 0.115 FPL in County ≥15% 476 44.4% 56.5% 24.6% 18.9% 43.5% Insurance Type HMO 188 17.5% 62.2% 21.3% 16.5% 0.005 37.8% 0.022 POS 742 69.2% 52.8% 25.1% 22.1% 47.2% PPO 142 13.2% 47.9% 19.0% 33.1% 52.1% INH Days Supply < 2 month 991 92.4% 54.5% 24.1% 21.4% 0.020 45.5% 0.126 Received on Date of 1st supply Fill ≥2 month 81 7.6% 45.7% 17.3% 37.0% 54.3% supply Year INH Regimen 2011 Q3–4 230 21.5% 58.3% 23.0% 18.7% 0.308 41.7% 0.298 Started 2012 Q1–4 450 42.0% 54.4% 21.8% 23.8% 45.6% 2013 Q1–4 346 32.3% 50.3% 26.3% 23.4% 49.7% 2014 Q1 46 4.3% 52.2% 23.9% 23.9% 47.8% State TB Rate – 3.85 3.84 3.81 0.846 3.83 0.864 LTBI Diagnostic Test TST 441 41.1% 53.5% 22.9% 23.6% < 0.001 46.5% 0.005 IGRA 219 20.4% 45.2% 23.7% 31.1% 54.8% Unknown/ 412 38.4% 58.7% 24.3% 17.0% 41.3% Other Percent Foreign Born in County – 19.96 20.24 20.97 0.403 20.60 0.516 Count of Clinical Risk None 662 61.8% 58.0% 22.2% 19.8% 0.011 42.0% 0.002 Factors 1 304 28.4% 47.7% 27.0% 25.3% 52.3% 2 or more 106 9.9% 45.3% 22.6% 32.1% 54.7% Stockbridge et al. BMC Public Health (2018) 18:662 Page 5 of 13 Table 2 Frequency distribution of patient characteristic variables for people initiating daily-dose isoniazid treatment and the proportion of people completing treatment by each characteristic. Treatment completion was categorized as 1) less than 6 months completed, 2) at least 6 months but less than 9 months completed, and 3) 9 or more months completed (Continued) Distribution % Achieving Each Level of Isoniazid Treatment % Completing ≥6 Mo. Completion N % or Mean < 6 Months ≥6 but < ≥9 Months p-value: 3 ≥6 Months p-value: < 6 of Total Complete 9 Months Complete Completion Complete vs [% or Mean] Complete [% or Levels [% or ≥6 Months [% or Mean] Mean] Complete Mean] Diagnosis of Contact w/ No 923 86.1% 54.3% 23.8% 21.9% 0.296 45.7% 0.457 TB diagnosis Had 149 13.9% 51.0% 22.2% 26.9% 49.0% diagnosis History of TB/ Late No 1027 95.8% 54.2% 23.1% 22.7% 0.426 45.8% 0.197 Effects diagnosis Had 45 4.2% 44.4% 35.6% 20.0% 55.6% diagnosis HIV Positive No 1030 96.1% 54.7% 23.4% 21.9% 0.004 45.3% 0.007 diagnosis Had 42 3.9% 33.3% 28.6% 38.1% 66.7% diagnosis Diabetes No 999 93.2% 54.5% 23.5% 22.0% 0.085 45.6% 0.126 diagnosis Had 73 6.8% 45.2% 24.7% 30.1% 54.8% diagnosis Tobacco No 1004 93.7% 54.2% 23.7% 22.1% 0.237 45.8% 0.366 diagnosis or medication Had 68 6.3% 48.5% 22.1% 29.4% 51.5% diagnosis or medication Immuno-suppressive No 948 88.4% 55.1% 23.0% 21.9% 0.030 44.9% 0.025 Medication medication Had 124 11.6% 44.4% 28.2% 27.4% 55.6% medication Based on an ICD-9-CM code of V01.1. Abbreviations: INH isoniazid, FPL federal poverty level, TB tuberculosis, TST tuberculin skin test, IGRA interferon-gamma release assays, LTBI latent tuberculosis infection, HMO health maintenance organization, POS point of service, PPO preferred provider organization medication use (Fig. 1; see Additional file 2 for logistic re- Additional post hoc analyses indicated that 34.9% of gression model results). the individuals initiating LTBI treatment had no proced- The results of the sensitivity models examining ure or diagnostic code in the medical claims data specif- ≥5 months of treatment were quite similar to the ically indicating that an LTBI test occurred, although the primary analyses wherein completion was defined as majority of these individuals had a diagnosis of LTBI ≥6 months of treatment (see Additional file 3 for de- (Table 4). We also identified significant associations be- tailed sensitivity model results). All findings were tween LTBI diagnostic test type and our model’s ex- directionally identical and odds ratios were of similar planatory variables (Table 5). Diagnostic test type was magnitude. While most variables were consistent in significantly associated with age, region, patient location, terms of statistical significance, there were two ex- insurance plan type, year, clinical risk factor count, his- ceptions. Some age group and insurance type cat- tory of or late effects of TB, HIV, diabetes, tobacco use, egories that were significant in the primary analyses and immunosuppressive medication use. were not significant in the sensitivity analyses. How- ever, the p-values for these categories approached Discussion significance, ranging from p = 0.052 to p =0.072. We used commercial insurance claims data to identify Based on these results we concluded that the results important individual, clinical, and system factors associ- of our primary analyses were robust to variations in ated with the completion of LTBI treatment with isonia- the definition of treatment completion. zid. Most striking were significant associations between Stockbridge et al. BMC Public Health (2018) 18:662 Page 6 of 13 Table 3 Results of two multivariable generalized ordered logit models with partial proportional odds which examine associations between patient characteristics and the completion of daily-dose isoniazid treatment for latent tuberculosis infection (N= 1072) Model 1: Includes Count of Clinical Risk Factors Model 2: Includes Specific Clinical Risk Factors Independent Variables Adjusted Odds Ratio 95% Confidence p-value Adjusted Odds Ratio 95% Confidence p-value Interval Interval Sex Female 1.000 1.000 Male 1.085 0.855 1.378 0.501 1.045 0.818 1.335 0.724 Age Group 0–14 1.000 1.000 15–29 0.547 0.351 0.854 0.008 0.552 0.353 0.863 0.009 30–44 0.597 0.385 0.925 0.021 0.599 0.386 0.930 0.022 45–64 0.584 0.370 0.920 0.020 0.574 0.362 0.909 0.018 Census Region Northeast 1.000 1.000 Midwest 0.934 0.588 1.483 0.772 0.933 0.587 1.484 0.771 South 0.716 0.466 1.102 0.129 0.692 0.449 1.069 0.097 West 0.989 0.676 1.448 0.956 0.967 0.661 1.416 0.864 Patient Location Neither regimen completed vs. ≥6 months completed (completed 6 or 9 month regimen) Large central metro county 1.000 1.000 Large fringe metro county 0.600 0.414 0.868 0.007 0.592 0.408 0.858 0.006 Any smaller county 0.767 0.495 1.189 0.235 0.776 0.500 1.203 0.256 < 9 months completed (neither regimen or 6 month regimen completed) vs. ≥9 months completed Large central metro county 1.000 1.000 Large fringe metro county 0.800 0.537 1.193 0.275 0.791 0.530 1.182 0.253 Any smaller county 0.767 0.495 1.189 0.235 0.776 0.500 1.203 0.256 % of Households < 15% 1.000 1.000 Under FPL in ≥15% 0.628 0.469 0.841 0.002 0.609 0.454 0.817 0.001 County Insurance Type Neither regimen completed vs. ≥6 months completed (completed 6 or 9 month regimen) HMO 1.000 1.000 POS 1.434 0.981 2.097 0.063 1.513 1.032 2.218 0.034 PPO 1.817 1.147 2.878 0.011 1.864 1.174 2.961 0.008 < 9 months completed (neither regimen or 6 month regimen completed) vs. ≥9 months completed HMO 1.000 1.000 POS 1.434 0.981 2.097 0.063 1.513 1.032 2.218 0.034 PPO 2.840 1.745 4.622 < 0.001 2.921 1.789 4.767 < 0.001 Prescription Size Neither regimen completed vs. ≥6 months completed (completed 6 or 9 month regimen) < 2 month supply 1.000 1.000 ≥2 month supply 1.419 0.884 2.278 0.148 1.395 0.867 2.245 0.170 < 9 months completed (neither regimen or 6 month regimen completed) vs. ≥9 months completed < 2 month supply 1.000 1.000 ≥2 month supply 2.268 1.383 3.720 0.001 2.233 1.359 3.670 0.002 Year INH Regimen 2011 Q3–4 1.000 1.000 Started 2012 Q1–4 1.109 0.802 1.532 0.531 1.104 0.798 1.526 0.551 2013 Q1–4 1.268 0.906 1.774 0.167 1.261 0.901 1.766 0.177 2014 Q1 1.333 0.720 2.468 0.361 1.333 0.718 2.473 0.363 State TB Rate 0.905 0.793 1.033 0.138 0.913 0.800 1.042 0.178 LTBI Diagnostic TST 1.000 1.000 Test IGRA 1.255 0.897 1.757 0.185 1.171 0.829 1.653 0.371 Stockbridge et al. BMC Public Health (2018) 18:662 Page 7 of 13 Table 3 Results of two multivariable generalized ordered logit models with partial proportional odds which examine associations between patient characteristics and the completion of daily-dose isoniazid treatment for latent tuberculosis infection (N= 1072) (Continued) Model 1: Includes Count of Clinical Risk Factors Model 2: Includes Specific Clinical Risk Factors Independent Variables Adjusted Odds Ratio 95% Confidence p-value Adjusted Odds Ratio 95% Confidence p-value Interval Interval Unknown/Other 0.813 0.616 1.071 0.141 0.812 0.615 1.071 0.141 Percent Foreign Born in County 1.004 0.989 1.019 0.612 1.004 0.989 1.019 0.636 Count of Clinical None 1.000 Risk Factors 1 1.522 1.158 2.001 0.003 na na na na 2 or more 1.816 1.188 2.778 0.006 na na na na Diagnosis of No diagnosis na na na na 1.000 Contact w/ TB Had diagnosis na na na na 1.289 0.916 1.814 0.145 History of TB/Late No diagnosis na na na na 1.000 Effects Had diagnosis na na na na 1.152 0.655 2.027 0.624 HIV Positive No diagnosis na na na na 1.000 Had diagnosis na na na na 2.578 1.377 4.827 0.003 Diabetes No diagnosis or medication na na na na 1.000 Had diagnosis or medication na na na na 1.458 0.902 2.355 0.124 Tobacco No diagnosis or medication na na na na 1.000 Had diagnosis or medication na na na na 1.254 0.766 2.052 0.368 Immuno- No medication na na na na 1.000 suppressive Had medication na na na na 1.470 0.997 2.167 0.052 Medications Constraints for parallel lines were applied to all independent variables except patient location, insurance type, and isoniazid days supply received For both models, isoniazid treatment completion was categorized as 1) less than 6 months completed, 2) at least 6 months but less than 9 months completed, and 3) 9 or more months completed Abbreviations: INH isoniazid, FPL federal poverty level, TB tuberculosis, TST tuberculin skin test, IGRA interferon-gamma release assays, LTBI latent tuberculosis infection, HMO health maintenance organization, POS point of service, PPO preferred provider organization a patient’s insurance plan type and treatment comple- (Healthcare Effectiveness Data and Information Set tion, suggesting that benefit design is a potential means [HEDIS]) [34]. Health plans’ quality improvement activ- to modify patient behaviors and ultimately TB risk. ities often focus on improving HEDIS rates, as many HMO plans, the most tightly managed insurance design, states consider quality assurance requirements met if were associated with the lowest likelihood of completion; plans maintain NCQA accreditation [33] and plans are PPO plans, the least restrictive plans, were associated required to calculate HEDIS measures to attain and with the highest. Completion differences may be due to maintain accreditation [35]. differences in access or cost sharing, as such health plan Pharmacy benefit design and prescribing offer similar characteristics are associated with continued adherence opportunities to decrease TB risk through improved to other types of medications [32]. treatment completion. Individuals filling larger prescrip- The lower completion rates for HMO-insured individ- tions (≥ 2 months supply) had greater odds of complet- uals suggest a need for HMOs to monitor and conduct ing a 9-month regimen. Although we cannot be certain quality improvement initiatives that improve enrollees’ given data limitations, completion of the longer regimen LTBI treatment completion rates. Such activities would may be due to the use of mail order pharmacies with not be unusual – HMOs in most states are required to automatic refill programs. Many insurers disallow com- operate quality assurance programs that involve moni- munity pharmacies from providing a > 1-month supply toring and conducting activities to improve care pro- of a medication. However, enrollees may be able to use cesses and clinical outcomes, such as improving mail order pharmacies to receive up to a 90-day supply medication adherence rates [33]. As private sector LTBI [36], and mail order pharmacies are more likely to have treatment becomes more common, the National Com- automatic refill programs [37]. These programs address mittee for Quality Assurance (NCQA) should consider patient passivity and transportation barriers by mailing incorporating an LTBI treatment completion measure prescription refills at regular intervals. Thus, encour- into its standard set of quality performance measures aging patients to fill larger prescriptions and use Stockbridge et al. BMC Public Health (2018) 18:662 Page 8 of 13 Fig. 1 Bar chart depicting the importance of variables in predicting completion of ≥6 months of isoniazid treatment for latent tuberculosis infection (LTBI). Longer bars represent greater importance automated mail order programs may increase 9-month increasingly likely as the total number of clinical risk factors isoniazid completion rates so long as appropriate clinical increased. Nevertheless, there are opportunities to improve monitoring to avoid hepatotoxicity and other complica- completion in high-risk private sector patients, as nearly tions is ensured [21]. half of those with one clinical risk factor and 45.3% of those Our analysis suggests that private sector providers are with > 1 risk factor did not complete at least 6 months of likely sensitive to and communicating the importance of LTBI treatment. As shorter-course regimens (e.g., 3 months treatment completion for LTBI patients at high risk of ac- of weekly isoniazid and rifapentine; 4 months of daily ri- tive TB. Patients with serious known risk factors such as fampin) typically have higher completion rates [38, 39], the HIV and immunosuppressive medication use [27]are more use of these regimens would likely increase treatment com- likely to complete treatment than others, and immunosup- pletion rates. We also found that TST is much more likely pressive medication use is of particular importance in pre- to be used among young children than IGRA. This is con- dicting adherence. Correspondingly, completion was sistent with the CDC guidelines [40] and suggests that Table 4 Frequency distribution of evidence of latent tuberculosis infection (LTBI) testing occurring in the 6 months prior to LTBI treatment initiation with isoniazid (n= 1072) Broad Categorization Used N % 95% Detailed Categorization N % 95% in Statistical Models Confidence Confidence Interval Interval TST 441 41.1% 38.2% 41.1% TST procedure code only, or TST code temporally first 441 41.1% 38.2% 44.1% IGRA 219 20.4% 18.1% 23.0% IGRA procedure code only, or IGRA code temporally first 219 20.4% 18.1% 23.0% Other/Unknown 412 38.4% 35.6% 41.4% IGRA & TST procedure codes present on same day 2 0.2% 0.0% 0.7% Other test for MTB occurred based on procedure code (no TST 5 0.5% 0.2% 1.1% or IGRA code) No procedure code provided information about testing, but a 31 2.9% 2.0% 4.1% diagnosis code indicated that screening occurred No procedure code or diagnosis code regarding testing was 261 24.4% 21.9% 27.0% present, but an LTBI diagnosis code was present Neither LTBI testing procedure nor diagnosis information 113 10.5% 8.8% 12.3% regarding LTBI was present Stockbridge et al. BMC Public Health (2018) 18:662 Page 9 of 13 Table 5 Bivariate associations between Mycobacterium tuberculosis test type and other patient characteristics. Includes people initiating daily-dose isoniazid treatment (N= 1072) Mycobacterium tuberculosis Test Type Tuberculin Skin Test Interferon-Gamma Other/ Unknown Test p-value [% or Mean] Release Assay [% or Mean] [% or Mean] Sex Female 42.1% 19.8% 38.1% 0.767 Male 40.0% 21.1% 38.8% Age Group 0–14 75.2% 8.6% 16.1% < 0.001 15–29 51.5% 11.0% 37.5% 30–44 36.1% 20.9% 43.0% 45–64 27.0% 31.3% 41.7% Census Region Northeast 46.6% 12.8% 40.6% 0.001 Midwest 36.8% 21.3% 41.9% South 41.9% 21.6% 36.5% West 37.9% 26.4% 35.7% Patient Location Large central metro 41.1% 23.4% 35.5% 0.033 county Large fringe metro 44.1% 17.2% 38.7% county Any smaller county 34.3% 20.0% 45.7% % of Households Under FPL in < 15% 41.9% 20.8% 37.3% 0.672 County ≥15% 40.1% 20.0% 39.9% Insurance Type HMO 38.8% 13.3% 47.9% 0.015 POS 41.1% 22.5% 36.4% PPO 44.4% 19.0% 36.6% Prescription Size < 2 month supply 41.5% 20.0% 38.5% 0.428 ≥2 month supply 37.0% 25.9% 37.0% Year INH Regimen Started 2011 Q3–4 49.1% 23.2% 38.7% 0.001 2012 Q1–4 36.2% 21.8% 42.0% 2013 Q1–4 40.5% 24.9% 34.7% 2014 Q1 54.4% 15.2% 30.4% State TB Rate 3.9 3.9 3.8 0.363 Percent Foreign Born in County 21.1 20.5 19.2 0.058 Count of Clinical Risk Factors None 46.8% 14.5% 38.7% < 0.001 1 36.8% 26.0% 37.2% 2 or more 17.9% 41.5% 40.6% Diagnosis of Contact w/ TB No diagnosis 39.8% 20.6% 39.6% 0.058 Had diagnosis 49.7% 19.5% 30.9% History of TB/Late Effects No diagnosis 42.0% 20.2% 37.9% 0.031 Had diagnosis 22.2% 36.7% 51.1% HIV No diagnosis 42.4% 19.0% 36.5% < 0.001 Had diagnosis 9.5% 54.8% 35.7% Diabetes No diagnosis or 42.3% 19.8% 37.8% 0.010 medication Had diagnosis or 24.7% 28.8% 46.6% medication Stockbridge et al. BMC Public Health (2018) 18:662 Page 10 of 13 Table 5 Bivariate associations between Mycobacterium tuberculosis test type and other patient characteristics. Includes people initiating daily-dose isoniazid treatment (N= 1072) (Continued) Mycobacterium tuberculosis Test Type Tuberculin Skin Test Interferon-Gamma Other/ Unknown Test p-value [% or Mean] Release Assay [% or Mean] [% or Mean] Tobacco No diagnosis or 42.1% 19.6% 38.3% 0.011 medication Had diagnosis or 26.5% 32.3% 41.2% medication Immunosuppressive No medication 43.8% 16.7% 39.6% < 0.001 Medications Had medication 21.0% 49.2% 29.8% Abbreviations: INH isoniazid, FPL federal poverty level, TB tuberculosis, LTBI latent tuberculosis infection, HMO health maintenance organization, POS point of service, PPO preferred provider organization private providers are receiving CDC messaging related to these other variables, there is no significant association be- best practices [21] and are following these practices. tween the receipt of an IGRA and treatment completion. We found that likelihood of completing ≥6 months of It is unclear if the use of IGRA facilitates completion or if treatment varied by patient location, with individuals in IGRA testing is more common in patients with other large fringe metro counties (i.e., suburban counties [23]) characteristics associated with completion. having a lower likelihood of completion than those in Claims are a rich source of information about com- large central metro counties (i.e., counties containing an mercial insurance-covered LTBI treatment occurring inner-city [23]). These findings are in contrast to recent across the US, but they have limitations. These data gen- research examining chronic condition medication adher- erally accurately reflect diagnoses and treatment [17], ence for rural, suburban, and urban populations in but accuracy varies with the clarity of coding instruc- which no significant differences were found [41]. The tions and guidelines [46]. There is ambiguity in the diag- differing LTBI treatment completion rates that we iden- nostic and procedure coding for LTBI. For example, tified may be due to differences in provider familiarity providers may be using the “contact with or exposure to with LTBI treatment best practices. Increased provider tuberculosis” diagnosis code to represent LTBI status ra- awareness of best practices and more years of experience ther than known recent contacts. This might explain in- are associated with increasing provider adherence to best consistencies between our findings and prior reports of practices [42, 43]. As TB incidence is much higher in better completion rates among TB contacts [47–50]. urban areas than other areas [44], providers in urban Conversely, many of our findings regarding LTBI treat- areas have likely had more exposure to patients in need ment completion are consistent with past research, in- of LTBI treatment, more exposure to LTBI treatment cluding associations with younger age and higher guidelines, and a greater awareness of the benefits of income [15, 16]. Additionally, claims data only reflect in- LTBI treatment completion. Claims data do not allow us formation submitted to a third party payer for the pur- to investigate providers’ knowledge of LTBI treatment poses of reimbursement [17]. Our finding that LTBI best practices, so additional research is warranted to testing procedure codes were not present in the claims confirm the cause of the location-related differences. for over a third of the individuals initiating isoniazid Even so, given the suburbanization of the US population treatment suggests that some providers are either not [45] and the importance of this variable in identifying billing for LTBI testing or some patients are receiving patients likely to complete < 6 months of treatment (see LTBI testing and treatment in different settings. For ex- Fig. 1), our findings identify an important opportunity to ample, a patient might be diagnosed for LTBI in a work- improve LTBI treatment completion rates in patients place, school, or public health department that does not treated by private sector providers in suburban areas. bill third party payers but subsequently seek treatment Our finding that IGRA is associated with greater likeli- or fill prescriptions in the private sector using insurance hood of treatment completion aligns with anecdotal re- benefits. ports that IGRA testing may yield greater diagnostic Due to limitations of claims data we cannot precisely de- confidence for patients and providers relative to TST. termine treatment intent or adherence, and conclusions However, the association is only significant in our un- about provider and patient behavior are based on inference, adjusted analysis. LTBI test type is also associated with not direct report. For instance, it is unclear whether a 6 or many other variables, including clinical risk factors, census 9-month treatment regimen was prescribed for a given pa- region, insurance plan type, and year. After adjusting for tient. Further, we cannot know if a filled prescription is Stockbridge et al. BMC Public Health (2018) 18:662 Page 11 of 13 actually consumed, and it is possible that those enrolled in Abbreviations CDC: Centers for Disease Control and Prevention; FPL: Federal poverty level; automatic refill programs may receive refills even if they HEDIS: Healthcare Effectiveness Data and Information Set; HIV: Human have discontinued their treatment. Of course, the uncer- immunodeficiency virus; HMO: Health maintenance organization; tainty related to medication consumption applies to all IGRA: Interferon-gamma release assays; INH: Isoniazid; LTBI: Latent tuberculosis infection; NCQA: National Committee for Quality Assurance; medication adherence research not involving direct obser- POS: Point of service; PPO: Preferred provider organization; TB: Tuberculosis; vation [51]. Fortunately, numerous studies have illustrated TST: Tuberculin skin test; US: United States; USPSTF: United States Preventive that medication adherence as measured by filled prescrip- Services Task Force tionsissignificantlycorrelated with both medication con- Acknowledgements sumption and drug serum levels [52]. Consequently, The authors gratefully acknowledge the support of the US Centers for claims-based methods of evaluating medication adherence Disease Control and Prevention’s Division of Tuberculosis Elimination and its are widely used in health services research and quality as- Tuberculosis Epidemiologic Studies Consortium (Atlanta, GA, USA) which provided valuable intellectual and other contributions. Additionally, the surance monitoring [53–62]. research reported in this publication was developed in collaboration with Data limitations left us unable to identify important TB Magellan Health, Inc. (Scottsdale, AZ, USA). We thank Magellan for their risk factors. Patient-level income and country of birth were invaluable contributions to this work. The findings and conclusions in this report are those of the authors and do not necessarily represent the official unavailable. While 59% of foreign-born people in the US position of the United States Centers for Disease Control and Prevention have private health insurance [13], claims data do not iden- (CDC) or Magellan Health, Inc. Mention of company names or products does tify nativity. However, county-level nativity and FPL rates not imply endorsement by the CDC or Magellan. were included as proxies. Our data also did not detail Funding treatment-related out-of-pocket costs for isoniazid or office No funding was received for this study. Dr. Stockbridge is a contractor for a visits, nor did it provide insight into insurance benefit plan commercial company: Magellan Health, Inc. Magellan Health provided design or network adequacy. Our analysis examining the support in the form of salaries for Dr. Stockbridge and access to the data, but did not have any additional role in the study design, analysis, decision to importance of the variables in the model should be inter- publish, or preparation of the manuscript. No other authors have financial preted with these limitations in mind, as the results only as- disclosures to report. sess the relative importance of variables available within the administrative claims data. Other, unavailable variables may Availability of data and materials be of great importance in predicting treatment completion. The data used in this study were licensed from Optum by Magellan Health, Inc. These data cannot be made freely available due to the nature of the Nevertheless, claims data provide unique opportunities to data (specifically, it contains dates related to individuals and their healthcare better understand LTBI treatment occurring in a setting of utilization) and due to the licensing agreement between Optum and increasing importance for TB prevention in the US. Magellan. Researchers interested in obtaining these data may contact Mike Crowley at Optum (mike.crowley@optum.com) in order to request clearance to use the data and to obtain a license for use of the data. Conclusions In the US, patient risks, provider and patient incentives or Authors’ contributions barriers, benefits design, and care processes in private ELS conceptualized the project, designed the methods, conducted data transformations and analyses, interpreted the results, drafted the manuscript, healthcare differ substantially from that of public health and approved the final version of the manuscript. TLM conceptualized the programs. Our findings illustrate that many of these factors project, designed the methods, interpreted the results, drafted the article, have an impact on LTBI treatment completion. This new and approved the final version. EKC contributed to the methodology design, interpreted results, revised the article, and approved the final version. CH information enables the development of evidence-based designed the methods, reviewed and approved the billing code lists, LTBI private sector treatment strategies. Such work is crit- interpreted results, revised the article, and approved the final version. ical as more private healthcare providers provide LTBI treatment and as public health authorities consider the op- Ethics approval and consent to participate The institutional review board of the University of North Texas Health portunities and limitations of private healthcare as a partner Science Center approved this project as exempt category research. The data to US TB elimination efforts. analyzed in the study consisted of medical and pharmacy claims data collected for non-research purposes. The data were de-identified and fully compliant with the US Health Insurance Portability and Accountability Act of Additional files 1996. This research did not involve the collection, use, or transmittal of indi- vidually identifiable data. Additional file 1: Excel file detailing the billing codes used in the analyses. Each tab provides information about a different variable. (XLS 385 kb) Competing interests The authors have no competing interests to declare. Dr. Stockbridge is a Additional file 2: Results of a logistic regression model which examines contractor for a commercial company: Magellan Health, Inc. This affiliation associations between patient characteristics and the completion of does not represent a competing interest and does not alter the authors’ ≥6 months of daily-dose isoniazid treatment for latent tuberculosis infec- adherence to BMC Public Health publication policies. tion (N = 1072). (XLSX 11 kb) Additional file 3: Results of logistic regression models which examine associations between patient characteristics and the completion of at Publisher’sNote least 5 months of daily-dose isoniazid treatment for latent tuberculosis in- Springer Nature remains neutral with regard to jurisdictional claims in fection (N = 1072). (XLS 31 kb) published maps and institutional affiliations. Stockbridge et al. BMC Public Health (2018) 18:662 Page 12 of 13 Author details treatment for latent tuberculosis infection: a systematic review. BMC Infect Department of Health Behavior and Health Systems, University of North Dis. 2016;16:257. Texas Health Science Center School of Public Health, 3500 Camp Bowie Blvd, 17. Virnig BA, McBean M. Administrative data for public health surveillance and Fort Worth, TX 76107, USA. Department of Advanced Health Analytics and planning. Annu Rev Public Health. 2001;22(1):213–30. Solutions, Magellan Health, Inc., 4800 N Scottsdale Rd #4400, Scottsdale, AZ 18. Optum: Clinformatics Data Mart Database. 2014. https://www.optum.com/ 85251, USA. Institute for Patient Safety, University of North Texas Health content/dam/optum/resources/productSheets/Clinformatics_for_Data_Mart.pdf. Science Center, 3500 Camp Bowie Blvd, Fort Worth, TX 76107, USA. College 19. Stockbridge EL, Miller TL, Carlson EK, Ho CS. Tuberculosis prevention in the of Nursing and Health Innovation, University of Texas at Arlington, 411 S. private sector: using claims-based methods to identify and evaluate latent Nedderman Drive, Arlington, TX 76019-0407, USA. Division of Tuberculosis tuberculosis infection treatment with isoniazid among the commercially insured. Elimination, Centers for Disease Control and Prevention, 1600 Clifton Road, J Public Health Manag Pract. 2017. Epub Ahead of Print Atlanta, GA 30333, USA. 20. Horsburgh CR Jr, Goldberg S, Bethel J, Chen S, Colson PW, Hirsch- Moverman Y, Hughes S, Shrestha-Kuwahara R, Sterling TR, Wall K, et al. Received: 14 January 2018 Accepted: 21 May 2018 Latent TB infection treatment acceptance and completion in the United States and Canada. Chest. 2010;137(2):401–9. 21. Centers for Disease Control and Prevention: Latent Tuberculosis Infection: A Guide For Primary Health Care Providers. 2013. http://www.cdc.gov/tb/ References publications/ltbi/pdf/targetedltbi.pdf. 1. Mancuso JD, Diffenderfer JM, Ghassemieh BJ, Horne DJ, Kao TC. The 22. Comstock GW. How much isoniazid is needed for prevention of prevalence of latent tuberculosis infection in the United States. Am J Respir tuberculosis among immunocompetent adults? Int J Tuberc Lung Dis. Crit Care Med. 2016;194(4):501–9. 1999;3(10):847–50. 2. Miramontes R, Hill AN, Yelk Woodruff RS, Lambert LA, Navin TR, Castro KG, 23. Centers for Disease Control and Prevention: NCHS Urban-Rural Classification LoBue PA. Tuberculosis infection in the United States: prevalence estimates Scheme for Counties. 2017. http://www.cdc.gov/nchs/data_access/urban_rural. from the National Health and nutrition examination survey, 2011-2012. PLoS htm. One. 2015;10(11):e0140881. 24. United States Census Bureau: Small Area Income and Poverty Estimates. 3. Kahwati LC, Feltner C, Halpern M, Woodell CL, Boland E, Amick HR, Palmieri 2015. https://www.census.gov/programs-surveys/saipe/data/datasets.html. Weber R, Jonas DE: Screening for Latent Tuberculosis Infection in Adults: An 25. United States Census Bureau: American Community Survey (ACS). 2018. Evidence Review for the U.S. Preventive Services Task Force. Evidence https://www.census.gov/programs-surveys/acs/. Synthesis No. 142. AHRQ Publication No. 14–05212-EF-1. Rockville, MD: 26. United States Department of Agriculture: Atlas of Rural and Small-Town Agency for Healthcare Research and Quality. In., vol. 14–05212-EF-1; 2016. America. 2018 http://www.ers.usda.gov/data-products/atlas-of-rural-and- 4. Centers for Disease Control and Prevention: Treatment regimens for latent small-town-america.aspx. TB infection (LTBI). 2017. https://www.cdc.gov/tb/topic/treatment/ltbi.htm. 27. Centers for Disease Control and Prevention. Targeted tuberculin testing and 5. Centers for Disease Control and Prevention: Division of Tuberculosis treatment of latent tuberculosis infection. American Thoracic Society. Elimination Strategic Plan 2016–2020. 2015. http://www.cdc.gov/tb/about/ MMWR. 2000;49(RR-6):1–51. strategicplan.htm. 28. Williams R. Generalized ordered logit/partial proportional odds models for 6. US Preventive Services Task Force: Update on Methods: Estimating Certainty and ordinal dependent variables. Stata J. 2006;6(1):58–82. Magnitude of Net Benefit. 2017. http://www.uspreventiveservicestaskforce.org/ 29. Saltelli A, Tarantola S, Campolongo F, Ratto M. Sensitivity analysis in Page/Name/update-on-methods-estimating-certainty-and-magnitude-of-net- practice: A Guide to Assessing Scientific Methods. West Sussex, England: benefit. John Wiley & Sons Ltd; 2004. 7. US Preventive Services Task Force, Bibbins-Domingo K, Grossman DC, Curry 30. IBM Corporation: IBM SPSS modeler 17 algorithms guide. Chicago, IL: IBM SJ, Bauman L, Davidson KW, Epling JWJ, Garcia FA, Herzstein J, Kemper AR, software Group; 2015. et al. Screening for latent tuberculosis infection in adults: US preventive 31. StataCorp: Stata Statistical Software: Release 14. College Station, TX: services task force recommendation statement. JAMA. 2016;316(9):962–9. StataCorp LP; 2016. 8. Sterling TR, Bethel J, Goldberg S, Weinfurter P, Yun L, Horsburgh CR, 32. Horgan CM,Merrick EL,Stewart MT, Scholle SH, Shih S.Improving medication Tuberculosis Epidemiologic Studies C. The scope and impact of treatment management of depression in health plans. Psychiatr Serv. 2008;59(1):72–7. of latent tuberculosis infection in the United States and Canada. Am J 33. Alexander BD, Broccolo BM, Daniels AR, SM DV, Hutto CB, Colborn-Loepere Respir Crit Care Med. 2006;173(8):927–31. C, Matyas DE, Mayo TW, Miles JJ, Polikoff EA, et al. Fundamentals of Health 9. Ehman M, Flood J, Barry PM. Tuberculosis treatment managed by providers Law, vol. 5th. Washington, DC: American Health Laywers Association; 2011. outside the public health department: lessons for the affordable care act. 34. National Committee for Quality Assurance: HEDIS & Performance Measures. PLoS One. 2014;9(10):e110645. 2018. http://www.ncqa.org/hedis-quality-measurement. 10. Institute of Medicine Committee on the Elimination of Tuberculosis in the 35. National Committee for Quality Assurance: Health Plan Accreditation. 2017. United States. Ending neglect: the elimination of tuberculosis in the U.S. http://www.ncqa.org/Programs/Accreditation/Health-Plan-HP.aspx. Washington (DC): national academy of sciences; 2000. 36. Pharmacy Benefits Management Institute: 2015–2016 Prescription Drug 11. Balaban V, Marks SM, Etkind SC, Katz DJ, Higashi J, Flood J, Cronin A, Ho CS, Benefit Cost and Plan Design Report. 2017. https://www.pbmi.com/ Khan A, Chorba T. Tuberculosis elimination efforts in the United States in ItemDetail?iProductCode=2015-2016_BDR&Category=BDR. the era of insurance expansion and the affordable care act. Public Health 37. Brown TA, Rickles NM. Mail-order versus local pharmacies on adherence: Rep. 2015;130(4):349–54. study methods make for unfair comparison. Am J Manag Care. 2010;16(3): 12. The Patient Protection and Affordable Care Act, Sec. 2713, Coverage of 226–7. author reply 227-228 Preventive Services. 2010. https://www.gpo.gov/fdsys/pkg/BILLS- 38. Sandgren AA, Noordegraaf-Schouten MV, van Kessel F, Stuurman A, 111hr3590enr/pdf/BILLS-111hr3590enr.pdf. Oordt-Speets A, van der Werf MJ. Initiation and completion rates for 13. Barnett JCV, Marina S: Health insurance coverage in the United States. 2016 latent tuberculosis infection treatment: a systematic review. BMC http://www.census.gov/content/dam/Census/library/publications/2016/ Infect Dis. 2016;16(1):204. demo/p60-257.pdf. 14. Foutz J, Damico A, Squires E, Garfield R. The Uninsured: A Primer - Key Facts 39. Sharma SK, Sharma A, Kadhiravan T, Tharyan P. Rifamycins (rifampicin, about Health Insurance and the Uninsured Under the Affordable Care Act. rifabutin and rifapentine) compared to isoniazid for preventing tuberculosis The Henry J. Kaiser Family Foundation. 2017.https://www.kff.org/uninsured/ in HIV-negative people at risk of active TB. Cochrane Database Syst Rev. report/the-uninsured-a-primer-key-facts-about-health-insurance-and-the- 2013;7:Cd007545. uninsured-under-the-affordable-care-act/. 40. Lewinsohn DM, Leonard MK, LoBue PA, et al. Official American Thoracic 15. Hirsch-Moverman Y, Daftary A, Franks J, Colson PW. Adherence to treatment for Society/Infectious Diseases Society of America/Centers for Disease Control latent tuberculosis infection: systematic review of studies in the US and Canada. and Prevention clinical practice guidelines: diagnosis of tuberculosis in The international journal of tuberculosis and lung disease. 12(11):1235–54. adults and children. Clin Infect Dis. 2017;64(2):e1-e33. 16. Stuurman AL, Noordegraaf-Schouten MV, Van Kessel F, Oordt-Speets A, 41. Arbuckle C, Tomaszewski D, Aronson BD, Brown L, Schommer J, Morisky Sandgren AA, Van der Werf MJ. Interventions for improving adherence to D, Linstead E. Evaluating factors impacting medication adherence Stockbridge et al. BMC Public Health (2018) 18:662 Page 13 of 13 among rural, urban, and suburban populations. J Rural Health. 2018; 58. Seabury SA, Lakdawalla DN, Dougherty JS, Sullivan J, Goldman DP. https://doi.org/10.1111/jrh.12291. Medication adherence and measures of health plan quality. Am J Manag 42. Francke AL, Smit MC, de Veer AJE, Mistiaen P. Factors influencing the Care. 2015;21(6):e379–89. implementation of clinical guidelines for health care professionals: a systematic 59. Kirkman MS, Rowan-Martin MT, Levin R, Fonseca VA, Schmittdiel JA, Herman meta-review. BMC Medical Informatics and Decision Making. 2008;8(38) WH, Aubert RE. Determinants of adherence to diabetes medications: findings from a large pharmacy claims database. Diabetes Care. 2015;38(4):604–9. 43. McKinlay JB, Link CL, Freund KM, Marceau LD, O’Donnell AB, Lutfey KL. 60. Doshi JA, Lim R, Li P, Young PP, Lawnicki VF, Troxel AB, Volpp KG. Sources of variation in physician adherence with clinical guidelines: results Synchronized prescription refills and medication adherence: a retrospective from a factorial experiment. J Gen Intern Med. 2007;22(3):289–96. claims analysis. Am J Manag Care. 2017;23(2):98–104. 44. Oren E, Winston CA, Pratt R, Robison VA, Narita M. Epidemiology of urban 61. Medicare Payment Advisory Commission: Chapter 7: Measuring the effects tuberculosis in the United States, 2000-2007. Am J Public Health. 2011; of medication adherence for the Medicare population. Report to the 101(7):1256–63. Congress: Medicare and the Health Care Delivery System. 2014. http://www. 45. Gardner T: Urban-suburban migration in the United States, 1955–2000. US medpac.gov/-documents-/reports. Census Bureau Center for Economic Studies Paper No. CES-WP-16-08: 2016. 62. Andrade SE, Kahler KH, Frech F, Chan KA. Methods for evaluation of https://doi.org/10.2139/ssrn.2726211. medication adherence and persistence using automated databases. 46. Centers for Disease Control and Prevention: Official Authorized Addenda: Pharmacoepidemiol Drug Saf. 2006;15(8):565–74. discussion 575-567 Human Immunodeficiency Virus Infection Codes and Official Guidelines for Coding and Reporting ICD-9-CM. 1994. https://wonder.cdc.gov/wonder/ prevguid/m0032908/m0032908.asp. 47. Li JJ. Adherence to treatment of latent tuberculosis infection in a clinical population in new York City. Int J Infect Dis. 2010;14(4):e292–7. 48. Nyamathi AM, Christiani A, Nahid P, Gregerson P, Leake B. A randomized controlled trial of two treatment programs for homeless adults with latent tuberculosis infection. Int J Tuberc Lung Dis. 2006;10(7):775–82. 49. Fiske CT, Yan FX, Hirsch-Moverman Y, Sterling TR, Reichler MR. Risk factors for treatment default in close contacts with latent tuberculous infection. Int J Tuberc Lung Dis. 2014;18(4):421–7. 50. Young KH, Ehman M, Reves R, Peterson BL, Khan A, Chorba TL, Jereb J. Tuberculosis contact investigations–United States, 2003-2012. MMWR Morb Mortal Wkly Rep. 2016;64(50–51):1369–74. 51. McGinnis B, Kauffman Y, Olson KL, Witt DM, Raebel MA. Interventions aimed at improving performance on medication adherence metrics. Int J Clin Pharm. 2014;36(1):20–5. 52. Steiner JF, Prochazka AV. The assessment of refill compliance using pharmacy records: methods, validity, and applications. J Clin Epidemiol. 1997;50(1):105–16. 53. Agency for Healthcare Research and Quality: Measure Summary: Medication management for people with asthma (NQMC: 010514). 2015. https://www.qualitymeasures.ahrq.gov/summaries/summary/49707/ medication-management-for-people-with-asthma-percentage-of- members-5-to-85-years-of-age-during-the-measurement-year-who-were- identified-as-having-persistent-asthma-and-who-were-dispensed-an- asthma-controller-medication-that-they-remained-on-for-at-least-75-of-?q= medication+management+for+people+with+asthma. 54. Agency for Healthcare Research and Quality: Measure Summary: Antidepressant medication management (effective acute phase treatment) (NQMC: 010531). 2015. https://www.qualitymeasures.ahrq.gov/summaries/ summary/49729/antidepressant-medication-management-effective-acute- phase-treatment-percentage-of-members-18-years-of-age-and-older-who- were-treated-with-antidepressant-medication-had-a-diagnosis-of-major- depression-and-who-remained-on-an-antidepressant-medication-for-at- lea?q=antidepressant+medication+management. 55. Agency for Healthcare Research and Quality: Measure Summary: Antidepressant medication management (effective continuation phase treatment) (NQMC: 010532). 2015. https://www.qualitymeasures.ahrq. gov/summaries/summary/49730/antidepressant-medication-management- effective-continuation-phase-treatment-percentage-of-members-18-years- of-age-and-older-who-were-treated-with-antidepressant-medication-had- a-diagnosis-of-major-depression-and-who-remained-on-an-antidepressant- medication-for?q=antidepressant+medication+management. 56. Agency for Healthcare Research and Quality: Measure Summary: Adherence to antipsychotic medications for individuals with schizophrenia (NQMC: 010540). 2015. https://www.qualitymeasures. ahrq.gov/summaries/summary/49738/adherence-to-antipsychotic- medications-for-individuals-with-schizophrenia-percentage-of-members- 19-to-64-years-of-age-during-the-measurement-year-with- schizophrenia-who-were-dispensed-and-remained-on-an-antipsychotic- t-of-their. medication-for-at-least-80-percen 57. Sueta CA, Rodgers JE, Chang PP, Zhou L, Thudium EM, Kucharska-Newton AM, Stearns SC. Medication adherence based on part D claims for patients with heart failure after hospitalization (from the atherosclerosis risk in communities study). Am J Cardiol. 2015;116(3):413–9. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BMC Public Health Springer Journals

Predictors of latent tuberculosis infection treatment completion in the US private sector: an analysis of administrative claims data

Free
13 pages
Loading next page...
 
/lp/springer_journal/predictors-of-latent-tuberculosis-infection-treatment-completion-in-pkH05uEKp1
Publisher
BioMed Central
Copyright
Copyright © 2018 by The Author(s).
Subject
Medicine & Public Health; Public Health; Medicine/Public Health, general; Epidemiology; Environmental Health; Biostatistics; Vaccine
eISSN
1471-2458
D.O.I.
10.1186/s12889-018-5578-3
Publisher site
See Article on Publisher Site

Abstract

Background: Factors that affect latent tuberculosis infection (LTBI) treatment completion in the US have not been well studied beyond public health settings. This gap was highlighted by recent health insurance-related regulatory changes that are likely to increase LTBI treatment by private sector healthcare providers. We analyzed LTBI treatment completion in the private healthcare setting to facilitate planning around this important opportunity for tuberculosis (TB) control in the US. Methods: We analyzed a national sample of commercial insurance medical and pharmacy claims data for people ages 0 to 64 years who initiated daily dose isoniazid treatment between July 2011 and March 2014 and who had complete data. All individuals resided in the US. Factors associated with treatment completion were examined using multivariable generalized ordered logit models and bivariate Kruskal-Wallis tests or Spearman correlations. Results: We identified 1072 individuals with complete data who initiated isoniazid LTBI treatment. Treatment completion was significantly associated with less restrictive health insurance, age < 15 years, patient location, use of interferon-gamma release assays, non-poverty, HIV diagnosis, immunosuppressive drug therapy, and higher cumulative counts of clinical risk factors. Conclusions: Private sector healthcare claims data provide insights into LTBI treatment completion patterns and patient/provider behaviors. Such information is critical to understanding the opportunities and limitations of private healthcare in the US to support treatment completion as this sector’s role in protecting against and eliminating TB grows. Keywords: Latent tuberculosis infection, LTBI, Treatment completion, Claims data, Administrative data, Isoniazid, Epidemiology, Health service delivery, Public health practice, Medication adherence Background develop TB over time, with higher progression risk among Up to 13 million people in the US have latent tuberculosis immunocompromised persons [3]. Although LTBI treat- infection (LTBI) [1, 2]. These people are infected with ment does not eliminate the risk of progression to active Mycobacterium tuberculosis yet do not have active tubercu- TB, completion of a proven LTBI treatment regimen (e.g., losis (TB) disease; they are asymptomatic and cannot trans- 6 or 9 months of daily isoniazid, 4 months of daily rifam- mitTB.Withouttreatment5–10% of people with LTBI will pin, 12 doses of weekly isoniazid and rifapentine) dramat- ically decreases that risk [4]. The US’ strategic plan to eliminate domestic TB includes risk-targeted identification * Correspondence: Erica.Stockbridge@unthsc.edu Department of Health Behavior and Health Systems, University of North and treatment of people with LTBI [5]. This strategy is Texas Health Science Center School of Public Health, 3500 Camp Bowie Blvd, supported by the US Preventive Services Task Force’s Fort Worth, TX 76107, USA (USPSTF) recent “Grade B” rating for LTBI testing in Department of Advanced Health Analytics and Solutions, Magellan Health, Inc., 4800 N Scottsdale Rd #4400, Scottsdale, AZ 85251, USA high-risk populations, which indicates to primary care Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Stockbridge et al. BMC Public Health (2018) 18:662 Page 2 of 13 providers that targeted LTBI testing and treatment afford indicate whether the 6 or 9-month regimen was prescribed, moderate health benefit with little risk [6, 7]. we could determine how many doses of isoniazid were dis- Public health agencies have traditionally provided most pensed. Thus, we grouped isoniazid treatments into three TB control and prevention services in the US [8–11]. How- mutually exclusive ordinal categories: 1) non-completion ever, the USPSTF’s rating and current policy will likely drive (< 180 doses received within 9-months), 2) completion of increased involvement by private sector providers as health the 6-month regimen but not the 9-month regimen (180 to insurers are now required to cover TB/LTBI testing in 269 doses received within 9-months), or 3) completion of high-risk populations with no patient cost sharing [12]. At the 9-month regimen (≥ 270 doses received within the same time, the uninsured rate in the US is decreasing 12-months) [20]. These increasing levels of completion are [13] and health insurance coverage is associated with in- important because, while isoniazid treatment completion at creased use of primary and other private sector health care any duration does not necessarily imply LTBI cure, the risk [14]. These shifts present an opportunity to coordinate of progression to active TB decreases as the duration of iso- public/private approaches to TB prevention. Factors associ- niazid treatment increases [22]. ated with LTBI treatment completion are seldom studied outside of public health settings [15, 16]. Differences in pa- Explanatory variables tient risks, provider and patient incentives, and care pro- Explanatory variables were constructed from the medical cesses in the private sector suggest a need for more and pharmacy claims data (see Additional file 1 for details). information about the factors associated with treatment Socio-demographic variables included sex, age, census re- completion in this increasingly important arena. gion, and a patient location variable based on the National We used a national sample of commercial claims data Center for Health Statistics urban-rural classification [23]. to examine private sector LTBI treatment across the US The percentage of households living under the federal pov- as a step toward filling this gap. Insurance claims offer a erty level in a patient’s county served as a proxy for house- window into private healthcare practice patterns [17]. hold income [24]. Additional variables included insurance We aimed to use these data to identify factors associated type (health maintenance organization [HMO], point of with the completion of daily dose isoniazid LTBI treat- service [POS], or preferred provider organization [PPO]), ment in the private sector setting. prescription size (the supply of isoniazid received when the first prescription was filled; < 2 months or ≥2months), Methods year, and the type of LTBI diagnostic test received in the Data source and analytic sample 6 months before treatment initiation. Non-clinical variables We analyzed de-identified medical and pharmacy claims related to risk of LTBI or progression to active TB were in- from Optum Clinformatics® Data Mart (formerly called cluded, such as the state TB rate. While country of birth the National Research Database) which includes claims was unavailable, we included prevalence of foreign-born in- for approximately 30.6 million commercially insured indi- dividuals in the patient’s county as a proxy [25, 26]. Clinical viduals – about 19% of the commercially insured US risk factors included diabetes, tobacco use, HIV, immuno- population [18]. We analyzed data for a randomly selected suppressive medication use, contact with or exposure to sample of 4 million people who were ages 0 to 64 years. TB, and a history of or late effects of TB [27]. A simple Additional details about this sample are described else- count of each patient’s clinical risk factors represented cu- where [19]. We used a claims-based algorithm to identify mulative risk (i.e., 0, 1, or > 1 risk factor). individual 6 to 9 month daily dose isoniazid courses of treatment for LTBI [19], which have been the most com- Statistical analyses monly used LTBI treatment regimens [20]. We examined We calculated the proportion of individuals in each of treatment initiated between July 2011 and March 2014. In three categories of treatment completion (i.e., < 6 months, addition to requiring that data be available to determine if 6 to < 9 months, ≥9 months) and examined the bivariate treatment was completed (as specified in the algorithm) relationships between the explanatory variables and com- [19] we required non-missing socio-demographic variables pletion using Kruskal-Wallis tests and Spearman correla- (i.e., the percent of foreign-born in county, patient loca- tions. We explored the adjusted association between these tion category, percent of households in county living variables and treatment completion category using multi- under the federal poverty level (FPL), and state TB rate). variable generalized ordered logit models. Variables meet- ing the parallel-lines assumption were constrained to have Measures equal effects; the odds ratios for non-completion versus Outcome variable completing ≥6 months of treatment and those for com- The outcome of interest was completion of daily isoniazid pleting < 9 months of treatment versus ≥9 months of treatment for LTBI [21]. Patients maybeprescribeda 6or treatment were the same. Variables violating the assump- 9-month isoniazid regimen [4]. While our data do not tion were not constrained and consequently have different Stockbridge et al. BMC Public Health (2018) 18:662 Page 3 of 13 odds ratios for completion category comparisons [28]. We Table 1 Completion of daily-dose isoniazid treatment for latent tuberculosis infection. N = 1072 ran two multivariable generalized ordered logit models. In Model 1 we examined the relationship between completion Isoniazid Treatment Completion Number % of 95% Confidence Total Interval and cumulative risk. Model 2 explored the relationship be- Less than 6 months (Incomplete 577 53.82 50.82–56.79 tween completion and individual clinical risk factors. treatment) We also ran a multivariable logit model with completion At least 6 months 495 46.18 43.20–49.17 of ≥6 months of treatment as the outcome measure and ≥6 months but < 9 months 253 23.60 21.15–26.24 all predictors from the more detailed Model 2 as explana- tory variables. This logit model was used to examine the ≥9 months 242 22.57 20.17–25.18 reduction of variance in the treatment completion variable attributable to each predictor, which provided insight into the importance of the variables with respect to model pre- Tables 2 and 3 describe relationships between the ex- dictions of completing ≥6months oftreatment [29, 30]. planatory variables and the likelihood of treatment com- We conducted two sets of post hoc analyses. First, in pletion from bivariate analyses and multivariable models, order to assess the robustness of our findings we conducted respectively. Significant unadjusted non-clinical factors as- sensitivity analyses using variations on our treatment com- sociated with completion included younger age, PPO in- pletion outcomes measure. We ran four multivariable logis- surance, larger prescription size, and residing in a county tic regression models to explore characteristics associated with < 15% of households below FPL. Similarly, in the with completion of ≥5 months of treatment and compare multivariable models younger people (ages 0 to 14 years) the results to the characteristics associated with ≥6months had higher adjusted odds of treatment completion than of treatment in Models 1 and 2. Four models were used be- older people. Compared to people in large central metro- cause we had two sets of explanatory variables (see descrip- politan counties, those in large fringe metropolitan coun- tions of Models 1 and 2 above), and we defined completion ties had lower adjusted odds of completing ≥6monthsof two ways: 1) 150 doses in 9 months, and 2) 150 doses in treatment, although this association was not seen with 8 months. We explored the data using two definitions be- completing ≥9 months of treatment. Residing in a county cause we identified no previous studies or clinical practice with ≥15% of households below FPL was significantly as- guidelines defining a time period in which 150 doses sociated with lower adjusted odds of completion. Detailed (5 months) of isoniazid would be considered completed adjusted odds ratios for the associations described above treatment. are found in Table 3. Second, we explored our findings related to the LTBI Insurance type and prescription size were also signifi- testing variable. We ran a frequency distribution which cantly associated with completion. The adjusted odds of a contained additional details about the LTBI tests re- PPO-insured patient completing ≥6 months of treatment ceived. Additionally, to clarify differences between the were 1.8 to 1.9 times that of an HMO-insured patient, and results in our bivariate and multivariable analyses, we the odds of a PPO-insured patient completing ≥9months conducted post hoc bivariate analyses exploring the rela- were 2.8 to 2.9 times that of an HMO-insured patient. Lar- tionship between the explanatory variables and the type ger prescription size was associated with higher adjusted of LTBI diagnostic test using chi square tests for cat- odds of completing ≥9 months of treatment, although this egorical variables and ANOVAs for continuous variables. association was not seen for completing ≥6monthsof We used Stata 14.2 for most statistical testing [31]but treatment. used IBM SPSS Modeler 17 to complete the importance IGRA testing, HIV, and immunosuppressive medica- analysis [32]. All statistical testing was two-sided, and sig- tion use each had statistically significant bivariate associ- nificance was tested at p <.05. ations with treatment completion. In the multivariable model, people with HIV had an adjusted 2.5 times greater odds of an increased level of completion relative Results to those without. Additionally, both unadjusted and ad- Two (0.2%) of 1074 individuals identified with the justed likelihood of completion was significantly associ- algorithm as having initiated isoniazid LTBI treat- ated with cumulative clinical risk. Compared to people ment were excluded due to missing geographic vari- with no clinical risk factors, those with one risk factor ables. Of the remaining 1072 almost half (46.2%) had 1.5 times greater adjusted odds and those with more completed ≥6 months of treatment. The balance than one risk factor had 1.8 times greater adjusted odds (53.8%) initiated but did not complete the minimum of an increased level of treatment completion. The im- 6-months course. Roughly equal proportions com- portance analysis indicated that the most important vari- pleted ≥6 but < 9 months treatment or ≥9months able in predicting treatment of ≥6 months of treatment was (23.6 and 22.6% of all patients, respectively; Table 1). patient location, followed closelybyimmunosuppressive Stockbridge et al. BMC Public Health (2018) 18:662 Page 4 of 13 Table 2 Frequency distribution of patient characteristic variables for people initiating daily-dose isoniazid treatment and the proportion of people completing treatment by each characteristic. Treatment completion was categorized as 1) less than 6 months completed, 2) at least 6 months but less than 9 months completed, and 3) 9 or more months completed Distribution % Achieving Each Level of Isoniazid Treatment % Completing ≥6 Mo. Completion N % or Mean < 6 Months ≥6 but < ≥9 Months p-value: 3 ≥6 Months p-value: < 6 of Total Complete 9 Months Complete Completion Complete vs [% or Mean] Complete [% or Levels [% or ≥6 Months [% or Mean] Mean] Complete Mean] Sex Female 575 53.6% 55.8% 22.1% 22.1% 0.232 44.2% 0.158 Male 497 46.4% 51.5% 25.4% 23.1% 48.5% Age Group 0–14 105 9.8% 43.8% 24.8% 31.4% 0.019 56.2% 0.064 15–29 291 27.1% 58.8% 23.4% 17.9% 41.2% 30–44 321 29.9% 53.9% 25.2% 20.9% 46.1% 45–64 355 33.1% 52.7% 22.0% 25.3% 47.3% Census Region Northeast 352 32.8% 54.8% 20.5% 24.7% 0.148 45.2% 0.151 Midwest 174 16.2% 52.3% 25.3% 22.4% 47.7% South 148 13.8% 61.5% 22.3% 16.2% 38.5% West 398 37.1% 53.8% 23.6% 22.6% 46.2% Patient Location Large 484 45.1% 50.0% 26.7% 23.4% 0.169 50.0% 0.066 central metro county Large fringe 413 38.5% 57.6% 19.6% 22.8% 42.4% metro county Any smaller 175 16.3% 55.4% 24.6% 20.0% 44.6% county % of Households Under < 15% 596 55.6% 51.7% 22.8% 25.5% 0.035 48.3% 0.115 FPL in County ≥15% 476 44.4% 56.5% 24.6% 18.9% 43.5% Insurance Type HMO 188 17.5% 62.2% 21.3% 16.5% 0.005 37.8% 0.022 POS 742 69.2% 52.8% 25.1% 22.1% 47.2% PPO 142 13.2% 47.9% 19.0% 33.1% 52.1% INH Days Supply < 2 month 991 92.4% 54.5% 24.1% 21.4% 0.020 45.5% 0.126 Received on Date of 1st supply Fill ≥2 month 81 7.6% 45.7% 17.3% 37.0% 54.3% supply Year INH Regimen 2011 Q3–4 230 21.5% 58.3% 23.0% 18.7% 0.308 41.7% 0.298 Started 2012 Q1–4 450 42.0% 54.4% 21.8% 23.8% 45.6% 2013 Q1–4 346 32.3% 50.3% 26.3% 23.4% 49.7% 2014 Q1 46 4.3% 52.2% 23.9% 23.9% 47.8% State TB Rate – 3.85 3.84 3.81 0.846 3.83 0.864 LTBI Diagnostic Test TST 441 41.1% 53.5% 22.9% 23.6% < 0.001 46.5% 0.005 IGRA 219 20.4% 45.2% 23.7% 31.1% 54.8% Unknown/ 412 38.4% 58.7% 24.3% 17.0% 41.3% Other Percent Foreign Born in County – 19.96 20.24 20.97 0.403 20.60 0.516 Count of Clinical Risk None 662 61.8% 58.0% 22.2% 19.8% 0.011 42.0% 0.002 Factors 1 304 28.4% 47.7% 27.0% 25.3% 52.3% 2 or more 106 9.9% 45.3% 22.6% 32.1% 54.7% Stockbridge et al. BMC Public Health (2018) 18:662 Page 5 of 13 Table 2 Frequency distribution of patient characteristic variables for people initiating daily-dose isoniazid treatment and the proportion of people completing treatment by each characteristic. Treatment completion was categorized as 1) less than 6 months completed, 2) at least 6 months but less than 9 months completed, and 3) 9 or more months completed (Continued) Distribution % Achieving Each Level of Isoniazid Treatment % Completing ≥6 Mo. Completion N % or Mean < 6 Months ≥6 but < ≥9 Months p-value: 3 ≥6 Months p-value: < 6 of Total Complete 9 Months Complete Completion Complete vs [% or Mean] Complete [% or Levels [% or ≥6 Months [% or Mean] Mean] Complete Mean] Diagnosis of Contact w/ No 923 86.1% 54.3% 23.8% 21.9% 0.296 45.7% 0.457 TB diagnosis Had 149 13.9% 51.0% 22.2% 26.9% 49.0% diagnosis History of TB/ Late No 1027 95.8% 54.2% 23.1% 22.7% 0.426 45.8% 0.197 Effects diagnosis Had 45 4.2% 44.4% 35.6% 20.0% 55.6% diagnosis HIV Positive No 1030 96.1% 54.7% 23.4% 21.9% 0.004 45.3% 0.007 diagnosis Had 42 3.9% 33.3% 28.6% 38.1% 66.7% diagnosis Diabetes No 999 93.2% 54.5% 23.5% 22.0% 0.085 45.6% 0.126 diagnosis Had 73 6.8% 45.2% 24.7% 30.1% 54.8% diagnosis Tobacco No 1004 93.7% 54.2% 23.7% 22.1% 0.237 45.8% 0.366 diagnosis or medication Had 68 6.3% 48.5% 22.1% 29.4% 51.5% diagnosis or medication Immuno-suppressive No 948 88.4% 55.1% 23.0% 21.9% 0.030 44.9% 0.025 Medication medication Had 124 11.6% 44.4% 28.2% 27.4% 55.6% medication Based on an ICD-9-CM code of V01.1. Abbreviations: INH isoniazid, FPL federal poverty level, TB tuberculosis, TST tuberculin skin test, IGRA interferon-gamma release assays, LTBI latent tuberculosis infection, HMO health maintenance organization, POS point of service, PPO preferred provider organization medication use (Fig. 1; see Additional file 2 for logistic re- Additional post hoc analyses indicated that 34.9% of gression model results). the individuals initiating LTBI treatment had no proced- The results of the sensitivity models examining ure or diagnostic code in the medical claims data specif- ≥5 months of treatment were quite similar to the ically indicating that an LTBI test occurred, although the primary analyses wherein completion was defined as majority of these individuals had a diagnosis of LTBI ≥6 months of treatment (see Additional file 3 for de- (Table 4). We also identified significant associations be- tailed sensitivity model results). All findings were tween LTBI diagnostic test type and our model’s ex- directionally identical and odds ratios were of similar planatory variables (Table 5). Diagnostic test type was magnitude. While most variables were consistent in significantly associated with age, region, patient location, terms of statistical significance, there were two ex- insurance plan type, year, clinical risk factor count, his- ceptions. Some age group and insurance type cat- tory of or late effects of TB, HIV, diabetes, tobacco use, egories that were significant in the primary analyses and immunosuppressive medication use. were not significant in the sensitivity analyses. How- ever, the p-values for these categories approached Discussion significance, ranging from p = 0.052 to p =0.072. We used commercial insurance claims data to identify Based on these results we concluded that the results important individual, clinical, and system factors associ- of our primary analyses were robust to variations in ated with the completion of LTBI treatment with isonia- the definition of treatment completion. zid. Most striking were significant associations between Stockbridge et al. BMC Public Health (2018) 18:662 Page 6 of 13 Table 3 Results of two multivariable generalized ordered logit models with partial proportional odds which examine associations between patient characteristics and the completion of daily-dose isoniazid treatment for latent tuberculosis infection (N= 1072) Model 1: Includes Count of Clinical Risk Factors Model 2: Includes Specific Clinical Risk Factors Independent Variables Adjusted Odds Ratio 95% Confidence p-value Adjusted Odds Ratio 95% Confidence p-value Interval Interval Sex Female 1.000 1.000 Male 1.085 0.855 1.378 0.501 1.045 0.818 1.335 0.724 Age Group 0–14 1.000 1.000 15–29 0.547 0.351 0.854 0.008 0.552 0.353 0.863 0.009 30–44 0.597 0.385 0.925 0.021 0.599 0.386 0.930 0.022 45–64 0.584 0.370 0.920 0.020 0.574 0.362 0.909 0.018 Census Region Northeast 1.000 1.000 Midwest 0.934 0.588 1.483 0.772 0.933 0.587 1.484 0.771 South 0.716 0.466 1.102 0.129 0.692 0.449 1.069 0.097 West 0.989 0.676 1.448 0.956 0.967 0.661 1.416 0.864 Patient Location Neither regimen completed vs. ≥6 months completed (completed 6 or 9 month regimen) Large central metro county 1.000 1.000 Large fringe metro county 0.600 0.414 0.868 0.007 0.592 0.408 0.858 0.006 Any smaller county 0.767 0.495 1.189 0.235 0.776 0.500 1.203 0.256 < 9 months completed (neither regimen or 6 month regimen completed) vs. ≥9 months completed Large central metro county 1.000 1.000 Large fringe metro county 0.800 0.537 1.193 0.275 0.791 0.530 1.182 0.253 Any smaller county 0.767 0.495 1.189 0.235 0.776 0.500 1.203 0.256 % of Households < 15% 1.000 1.000 Under FPL in ≥15% 0.628 0.469 0.841 0.002 0.609 0.454 0.817 0.001 County Insurance Type Neither regimen completed vs. ≥6 months completed (completed 6 or 9 month regimen) HMO 1.000 1.000 POS 1.434 0.981 2.097 0.063 1.513 1.032 2.218 0.034 PPO 1.817 1.147 2.878 0.011 1.864 1.174 2.961 0.008 < 9 months completed (neither regimen or 6 month regimen completed) vs. ≥9 months completed HMO 1.000 1.000 POS 1.434 0.981 2.097 0.063 1.513 1.032 2.218 0.034 PPO 2.840 1.745 4.622 < 0.001 2.921 1.789 4.767 < 0.001 Prescription Size Neither regimen completed vs. ≥6 months completed (completed 6 or 9 month regimen) < 2 month supply 1.000 1.000 ≥2 month supply 1.419 0.884 2.278 0.148 1.395 0.867 2.245 0.170 < 9 months completed (neither regimen or 6 month regimen completed) vs. ≥9 months completed < 2 month supply 1.000 1.000 ≥2 month supply 2.268 1.383 3.720 0.001 2.233 1.359 3.670 0.002 Year INH Regimen 2011 Q3–4 1.000 1.000 Started 2012 Q1–4 1.109 0.802 1.532 0.531 1.104 0.798 1.526 0.551 2013 Q1–4 1.268 0.906 1.774 0.167 1.261 0.901 1.766 0.177 2014 Q1 1.333 0.720 2.468 0.361 1.333 0.718 2.473 0.363 State TB Rate 0.905 0.793 1.033 0.138 0.913 0.800 1.042 0.178 LTBI Diagnostic TST 1.000 1.000 Test IGRA 1.255 0.897 1.757 0.185 1.171 0.829 1.653 0.371 Stockbridge et al. BMC Public Health (2018) 18:662 Page 7 of 13 Table 3 Results of two multivariable generalized ordered logit models with partial proportional odds which examine associations between patient characteristics and the completion of daily-dose isoniazid treatment for latent tuberculosis infection (N= 1072) (Continued) Model 1: Includes Count of Clinical Risk Factors Model 2: Includes Specific Clinical Risk Factors Independent Variables Adjusted Odds Ratio 95% Confidence p-value Adjusted Odds Ratio 95% Confidence p-value Interval Interval Unknown/Other 0.813 0.616 1.071 0.141 0.812 0.615 1.071 0.141 Percent Foreign Born in County 1.004 0.989 1.019 0.612 1.004 0.989 1.019 0.636 Count of Clinical None 1.000 Risk Factors 1 1.522 1.158 2.001 0.003 na na na na 2 or more 1.816 1.188 2.778 0.006 na na na na Diagnosis of No diagnosis na na na na 1.000 Contact w/ TB Had diagnosis na na na na 1.289 0.916 1.814 0.145 History of TB/Late No diagnosis na na na na 1.000 Effects Had diagnosis na na na na 1.152 0.655 2.027 0.624 HIV Positive No diagnosis na na na na 1.000 Had diagnosis na na na na 2.578 1.377 4.827 0.003 Diabetes No diagnosis or medication na na na na 1.000 Had diagnosis or medication na na na na 1.458 0.902 2.355 0.124 Tobacco No diagnosis or medication na na na na 1.000 Had diagnosis or medication na na na na 1.254 0.766 2.052 0.368 Immuno- No medication na na na na 1.000 suppressive Had medication na na na na 1.470 0.997 2.167 0.052 Medications Constraints for parallel lines were applied to all independent variables except patient location, insurance type, and isoniazid days supply received For both models, isoniazid treatment completion was categorized as 1) less than 6 months completed, 2) at least 6 months but less than 9 months completed, and 3) 9 or more months completed Abbreviations: INH isoniazid, FPL federal poverty level, TB tuberculosis, TST tuberculin skin test, IGRA interferon-gamma release assays, LTBI latent tuberculosis infection, HMO health maintenance organization, POS point of service, PPO preferred provider organization a patient’s insurance plan type and treatment comple- (Healthcare Effectiveness Data and Information Set tion, suggesting that benefit design is a potential means [HEDIS]) [34]. Health plans’ quality improvement activ- to modify patient behaviors and ultimately TB risk. ities often focus on improving HEDIS rates, as many HMO plans, the most tightly managed insurance design, states consider quality assurance requirements met if were associated with the lowest likelihood of completion; plans maintain NCQA accreditation [33] and plans are PPO plans, the least restrictive plans, were associated required to calculate HEDIS measures to attain and with the highest. Completion differences may be due to maintain accreditation [35]. differences in access or cost sharing, as such health plan Pharmacy benefit design and prescribing offer similar characteristics are associated with continued adherence opportunities to decrease TB risk through improved to other types of medications [32]. treatment completion. Individuals filling larger prescrip- The lower completion rates for HMO-insured individ- tions (≥ 2 months supply) had greater odds of complet- uals suggest a need for HMOs to monitor and conduct ing a 9-month regimen. Although we cannot be certain quality improvement initiatives that improve enrollees’ given data limitations, completion of the longer regimen LTBI treatment completion rates. Such activities would may be due to the use of mail order pharmacies with not be unusual – HMOs in most states are required to automatic refill programs. Many insurers disallow com- operate quality assurance programs that involve moni- munity pharmacies from providing a > 1-month supply toring and conducting activities to improve care pro- of a medication. However, enrollees may be able to use cesses and clinical outcomes, such as improving mail order pharmacies to receive up to a 90-day supply medication adherence rates [33]. As private sector LTBI [36], and mail order pharmacies are more likely to have treatment becomes more common, the National Com- automatic refill programs [37]. These programs address mittee for Quality Assurance (NCQA) should consider patient passivity and transportation barriers by mailing incorporating an LTBI treatment completion measure prescription refills at regular intervals. Thus, encour- into its standard set of quality performance measures aging patients to fill larger prescriptions and use Stockbridge et al. BMC Public Health (2018) 18:662 Page 8 of 13 Fig. 1 Bar chart depicting the importance of variables in predicting completion of ≥6 months of isoniazid treatment for latent tuberculosis infection (LTBI). Longer bars represent greater importance automated mail order programs may increase 9-month increasingly likely as the total number of clinical risk factors isoniazid completion rates so long as appropriate clinical increased. Nevertheless, there are opportunities to improve monitoring to avoid hepatotoxicity and other complica- completion in high-risk private sector patients, as nearly tions is ensured [21]. half of those with one clinical risk factor and 45.3% of those Our analysis suggests that private sector providers are with > 1 risk factor did not complete at least 6 months of likely sensitive to and communicating the importance of LTBI treatment. As shorter-course regimens (e.g., 3 months treatment completion for LTBI patients at high risk of ac- of weekly isoniazid and rifapentine; 4 months of daily ri- tive TB. Patients with serious known risk factors such as fampin) typically have higher completion rates [38, 39], the HIV and immunosuppressive medication use [27]are more use of these regimens would likely increase treatment com- likely to complete treatment than others, and immunosup- pletion rates. We also found that TST is much more likely pressive medication use is of particular importance in pre- to be used among young children than IGRA. This is con- dicting adherence. Correspondingly, completion was sistent with the CDC guidelines [40] and suggests that Table 4 Frequency distribution of evidence of latent tuberculosis infection (LTBI) testing occurring in the 6 months prior to LTBI treatment initiation with isoniazid (n= 1072) Broad Categorization Used N % 95% Detailed Categorization N % 95% in Statistical Models Confidence Confidence Interval Interval TST 441 41.1% 38.2% 41.1% TST procedure code only, or TST code temporally first 441 41.1% 38.2% 44.1% IGRA 219 20.4% 18.1% 23.0% IGRA procedure code only, or IGRA code temporally first 219 20.4% 18.1% 23.0% Other/Unknown 412 38.4% 35.6% 41.4% IGRA & TST procedure codes present on same day 2 0.2% 0.0% 0.7% Other test for MTB occurred based on procedure code (no TST 5 0.5% 0.2% 1.1% or IGRA code) No procedure code provided information about testing, but a 31 2.9% 2.0% 4.1% diagnosis code indicated that screening occurred No procedure code or diagnosis code regarding testing was 261 24.4% 21.9% 27.0% present, but an LTBI diagnosis code was present Neither LTBI testing procedure nor diagnosis information 113 10.5% 8.8% 12.3% regarding LTBI was present Stockbridge et al. BMC Public Health (2018) 18:662 Page 9 of 13 Table 5 Bivariate associations between Mycobacterium tuberculosis test type and other patient characteristics. Includes people initiating daily-dose isoniazid treatment (N= 1072) Mycobacterium tuberculosis Test Type Tuberculin Skin Test Interferon-Gamma Other/ Unknown Test p-value [% or Mean] Release Assay [% or Mean] [% or Mean] Sex Female 42.1% 19.8% 38.1% 0.767 Male 40.0% 21.1% 38.8% Age Group 0–14 75.2% 8.6% 16.1% < 0.001 15–29 51.5% 11.0% 37.5% 30–44 36.1% 20.9% 43.0% 45–64 27.0% 31.3% 41.7% Census Region Northeast 46.6% 12.8% 40.6% 0.001 Midwest 36.8% 21.3% 41.9% South 41.9% 21.6% 36.5% West 37.9% 26.4% 35.7% Patient Location Large central metro 41.1% 23.4% 35.5% 0.033 county Large fringe metro 44.1% 17.2% 38.7% county Any smaller county 34.3% 20.0% 45.7% % of Households Under FPL in < 15% 41.9% 20.8% 37.3% 0.672 County ≥15% 40.1% 20.0% 39.9% Insurance Type HMO 38.8% 13.3% 47.9% 0.015 POS 41.1% 22.5% 36.4% PPO 44.4% 19.0% 36.6% Prescription Size < 2 month supply 41.5% 20.0% 38.5% 0.428 ≥2 month supply 37.0% 25.9% 37.0% Year INH Regimen Started 2011 Q3–4 49.1% 23.2% 38.7% 0.001 2012 Q1–4 36.2% 21.8% 42.0% 2013 Q1–4 40.5% 24.9% 34.7% 2014 Q1 54.4% 15.2% 30.4% State TB Rate 3.9 3.9 3.8 0.363 Percent Foreign Born in County 21.1 20.5 19.2 0.058 Count of Clinical Risk Factors None 46.8% 14.5% 38.7% < 0.001 1 36.8% 26.0% 37.2% 2 or more 17.9% 41.5% 40.6% Diagnosis of Contact w/ TB No diagnosis 39.8% 20.6% 39.6% 0.058 Had diagnosis 49.7% 19.5% 30.9% History of TB/Late Effects No diagnosis 42.0% 20.2% 37.9% 0.031 Had diagnosis 22.2% 36.7% 51.1% HIV No diagnosis 42.4% 19.0% 36.5% < 0.001 Had diagnosis 9.5% 54.8% 35.7% Diabetes No diagnosis or 42.3% 19.8% 37.8% 0.010 medication Had diagnosis or 24.7% 28.8% 46.6% medication Stockbridge et al. BMC Public Health (2018) 18:662 Page 10 of 13 Table 5 Bivariate associations between Mycobacterium tuberculosis test type and other patient characteristics. Includes people initiating daily-dose isoniazid treatment (N= 1072) (Continued) Mycobacterium tuberculosis Test Type Tuberculin Skin Test Interferon-Gamma Other/ Unknown Test p-value [% or Mean] Release Assay [% or Mean] [% or Mean] Tobacco No diagnosis or 42.1% 19.6% 38.3% 0.011 medication Had diagnosis or 26.5% 32.3% 41.2% medication Immunosuppressive No medication 43.8% 16.7% 39.6% < 0.001 Medications Had medication 21.0% 49.2% 29.8% Abbreviations: INH isoniazid, FPL federal poverty level, TB tuberculosis, LTBI latent tuberculosis infection, HMO health maintenance organization, POS point of service, PPO preferred provider organization private providers are receiving CDC messaging related to these other variables, there is no significant association be- best practices [21] and are following these practices. tween the receipt of an IGRA and treatment completion. We found that likelihood of completing ≥6 months of It is unclear if the use of IGRA facilitates completion or if treatment varied by patient location, with individuals in IGRA testing is more common in patients with other large fringe metro counties (i.e., suburban counties [23]) characteristics associated with completion. having a lower likelihood of completion than those in Claims are a rich source of information about com- large central metro counties (i.e., counties containing an mercial insurance-covered LTBI treatment occurring inner-city [23]). These findings are in contrast to recent across the US, but they have limitations. These data gen- research examining chronic condition medication adher- erally accurately reflect diagnoses and treatment [17], ence for rural, suburban, and urban populations in but accuracy varies with the clarity of coding instruc- which no significant differences were found [41]. The tions and guidelines [46]. There is ambiguity in the diag- differing LTBI treatment completion rates that we iden- nostic and procedure coding for LTBI. For example, tified may be due to differences in provider familiarity providers may be using the “contact with or exposure to with LTBI treatment best practices. Increased provider tuberculosis” diagnosis code to represent LTBI status ra- awareness of best practices and more years of experience ther than known recent contacts. This might explain in- are associated with increasing provider adherence to best consistencies between our findings and prior reports of practices [42, 43]. As TB incidence is much higher in better completion rates among TB contacts [47–50]. urban areas than other areas [44], providers in urban Conversely, many of our findings regarding LTBI treat- areas have likely had more exposure to patients in need ment completion are consistent with past research, in- of LTBI treatment, more exposure to LTBI treatment cluding associations with younger age and higher guidelines, and a greater awareness of the benefits of income [15, 16]. Additionally, claims data only reflect in- LTBI treatment completion. Claims data do not allow us formation submitted to a third party payer for the pur- to investigate providers’ knowledge of LTBI treatment poses of reimbursement [17]. Our finding that LTBI best practices, so additional research is warranted to testing procedure codes were not present in the claims confirm the cause of the location-related differences. for over a third of the individuals initiating isoniazid Even so, given the suburbanization of the US population treatment suggests that some providers are either not [45] and the importance of this variable in identifying billing for LTBI testing or some patients are receiving patients likely to complete < 6 months of treatment (see LTBI testing and treatment in different settings. For ex- Fig. 1), our findings identify an important opportunity to ample, a patient might be diagnosed for LTBI in a work- improve LTBI treatment completion rates in patients place, school, or public health department that does not treated by private sector providers in suburban areas. bill third party payers but subsequently seek treatment Our finding that IGRA is associated with greater likeli- or fill prescriptions in the private sector using insurance hood of treatment completion aligns with anecdotal re- benefits. ports that IGRA testing may yield greater diagnostic Due to limitations of claims data we cannot precisely de- confidence for patients and providers relative to TST. termine treatment intent or adherence, and conclusions However, the association is only significant in our un- about provider and patient behavior are based on inference, adjusted analysis. LTBI test type is also associated with not direct report. For instance, it is unclear whether a 6 or many other variables, including clinical risk factors, census 9-month treatment regimen was prescribed for a given pa- region, insurance plan type, and year. After adjusting for tient. Further, we cannot know if a filled prescription is Stockbridge et al. BMC Public Health (2018) 18:662 Page 11 of 13 actually consumed, and it is possible that those enrolled in Abbreviations CDC: Centers for Disease Control and Prevention; FPL: Federal poverty level; automatic refill programs may receive refills even if they HEDIS: Healthcare Effectiveness Data and Information Set; HIV: Human have discontinued their treatment. Of course, the uncer- immunodeficiency virus; HMO: Health maintenance organization; tainty related to medication consumption applies to all IGRA: Interferon-gamma release assays; INH: Isoniazid; LTBI: Latent tuberculosis infection; NCQA: National Committee for Quality Assurance; medication adherence research not involving direct obser- POS: Point of service; PPO: Preferred provider organization; TB: Tuberculosis; vation [51]. Fortunately, numerous studies have illustrated TST: Tuberculin skin test; US: United States; USPSTF: United States Preventive that medication adherence as measured by filled prescrip- Services Task Force tionsissignificantlycorrelated with both medication con- Acknowledgements sumption and drug serum levels [52]. Consequently, The authors gratefully acknowledge the support of the US Centers for claims-based methods of evaluating medication adherence Disease Control and Prevention’s Division of Tuberculosis Elimination and its are widely used in health services research and quality as- Tuberculosis Epidemiologic Studies Consortium (Atlanta, GA, USA) which provided valuable intellectual and other contributions. Additionally, the surance monitoring [53–62]. research reported in this publication was developed in collaboration with Data limitations left us unable to identify important TB Magellan Health, Inc. (Scottsdale, AZ, USA). We thank Magellan for their risk factors. Patient-level income and country of birth were invaluable contributions to this work. The findings and conclusions in this report are those of the authors and do not necessarily represent the official unavailable. While 59% of foreign-born people in the US position of the United States Centers for Disease Control and Prevention have private health insurance [13], claims data do not iden- (CDC) or Magellan Health, Inc. Mention of company names or products does tify nativity. However, county-level nativity and FPL rates not imply endorsement by the CDC or Magellan. were included as proxies. Our data also did not detail Funding treatment-related out-of-pocket costs for isoniazid or office No funding was received for this study. Dr. Stockbridge is a contractor for a visits, nor did it provide insight into insurance benefit plan commercial company: Magellan Health, Inc. Magellan Health provided design or network adequacy. Our analysis examining the support in the form of salaries for Dr. Stockbridge and access to the data, but did not have any additional role in the study design, analysis, decision to importance of the variables in the model should be inter- publish, or preparation of the manuscript. No other authors have financial preted with these limitations in mind, as the results only as- disclosures to report. sess the relative importance of variables available within the administrative claims data. Other, unavailable variables may Availability of data and materials be of great importance in predicting treatment completion. The data used in this study were licensed from Optum by Magellan Health, Inc. These data cannot be made freely available due to the nature of the Nevertheless, claims data provide unique opportunities to data (specifically, it contains dates related to individuals and their healthcare better understand LTBI treatment occurring in a setting of utilization) and due to the licensing agreement between Optum and increasing importance for TB prevention in the US. Magellan. Researchers interested in obtaining these data may contact Mike Crowley at Optum (mike.crowley@optum.com) in order to request clearance to use the data and to obtain a license for use of the data. Conclusions In the US, patient risks, provider and patient incentives or Authors’ contributions barriers, benefits design, and care processes in private ELS conceptualized the project, designed the methods, conducted data transformations and analyses, interpreted the results, drafted the manuscript, healthcare differ substantially from that of public health and approved the final version of the manuscript. TLM conceptualized the programs. Our findings illustrate that many of these factors project, designed the methods, interpreted the results, drafted the article, have an impact on LTBI treatment completion. This new and approved the final version. EKC contributed to the methodology design, interpreted results, revised the article, and approved the final version. CH information enables the development of evidence-based designed the methods, reviewed and approved the billing code lists, LTBI private sector treatment strategies. Such work is crit- interpreted results, revised the article, and approved the final version. ical as more private healthcare providers provide LTBI treatment and as public health authorities consider the op- Ethics approval and consent to participate The institutional review board of the University of North Texas Health portunities and limitations of private healthcare as a partner Science Center approved this project as exempt category research. The data to US TB elimination efforts. analyzed in the study consisted of medical and pharmacy claims data collected for non-research purposes. The data were de-identified and fully compliant with the US Health Insurance Portability and Accountability Act of Additional files 1996. This research did not involve the collection, use, or transmittal of indi- vidually identifiable data. Additional file 1: Excel file detailing the billing codes used in the analyses. Each tab provides information about a different variable. (XLS 385 kb) Competing interests The authors have no competing interests to declare. Dr. Stockbridge is a Additional file 2: Results of a logistic regression model which examines contractor for a commercial company: Magellan Health, Inc. This affiliation associations between patient characteristics and the completion of does not represent a competing interest and does not alter the authors’ ≥6 months of daily-dose isoniazid treatment for latent tuberculosis infec- adherence to BMC Public Health publication policies. tion (N = 1072). (XLSX 11 kb) Additional file 3: Results of logistic regression models which examine associations between patient characteristics and the completion of at Publisher’sNote least 5 months of daily-dose isoniazid treatment for latent tuberculosis in- Springer Nature remains neutral with regard to jurisdictional claims in fection (N = 1072). (XLS 31 kb) published maps and institutional affiliations. Stockbridge et al. BMC Public Health (2018) 18:662 Page 12 of 13 Author details treatment for latent tuberculosis infection: a systematic review. BMC Infect Department of Health Behavior and Health Systems, University of North Dis. 2016;16:257. Texas Health Science Center School of Public Health, 3500 Camp Bowie Blvd, 17. Virnig BA, McBean M. Administrative data for public health surveillance and Fort Worth, TX 76107, USA. Department of Advanced Health Analytics and planning. Annu Rev Public Health. 2001;22(1):213–30. Solutions, Magellan Health, Inc., 4800 N Scottsdale Rd #4400, Scottsdale, AZ 18. Optum: Clinformatics Data Mart Database. 2014. https://www.optum.com/ 85251, USA. Institute for Patient Safety, University of North Texas Health content/dam/optum/resources/productSheets/Clinformatics_for_Data_Mart.pdf. Science Center, 3500 Camp Bowie Blvd, Fort Worth, TX 76107, USA. College 19. Stockbridge EL, Miller TL, Carlson EK, Ho CS. Tuberculosis prevention in the of Nursing and Health Innovation, University of Texas at Arlington, 411 S. private sector: using claims-based methods to identify and evaluate latent Nedderman Drive, Arlington, TX 76019-0407, USA. Division of Tuberculosis tuberculosis infection treatment with isoniazid among the commercially insured. Elimination, Centers for Disease Control and Prevention, 1600 Clifton Road, J Public Health Manag Pract. 2017. Epub Ahead of Print Atlanta, GA 30333, USA. 20. Horsburgh CR Jr, Goldberg S, Bethel J, Chen S, Colson PW, Hirsch- Moverman Y, Hughes S, Shrestha-Kuwahara R, Sterling TR, Wall K, et al. Received: 14 January 2018 Accepted: 21 May 2018 Latent TB infection treatment acceptance and completion in the United States and Canada. Chest. 2010;137(2):401–9. 21. Centers for Disease Control and Prevention: Latent Tuberculosis Infection: A Guide For Primary Health Care Providers. 2013. http://www.cdc.gov/tb/ References publications/ltbi/pdf/targetedltbi.pdf. 1. Mancuso JD, Diffenderfer JM, Ghassemieh BJ, Horne DJ, Kao TC. The 22. Comstock GW. How much isoniazid is needed for prevention of prevalence of latent tuberculosis infection in the United States. Am J Respir tuberculosis among immunocompetent adults? Int J Tuberc Lung Dis. Crit Care Med. 2016;194(4):501–9. 1999;3(10):847–50. 2. Miramontes R, Hill AN, Yelk Woodruff RS, Lambert LA, Navin TR, Castro KG, 23. Centers for Disease Control and Prevention: NCHS Urban-Rural Classification LoBue PA. Tuberculosis infection in the United States: prevalence estimates Scheme for Counties. 2017. http://www.cdc.gov/nchs/data_access/urban_rural. from the National Health and nutrition examination survey, 2011-2012. PLoS htm. One. 2015;10(11):e0140881. 24. United States Census Bureau: Small Area Income and Poverty Estimates. 3. Kahwati LC, Feltner C, Halpern M, Woodell CL, Boland E, Amick HR, Palmieri 2015. https://www.census.gov/programs-surveys/saipe/data/datasets.html. Weber R, Jonas DE: Screening for Latent Tuberculosis Infection in Adults: An 25. United States Census Bureau: American Community Survey (ACS). 2018. Evidence Review for the U.S. Preventive Services Task Force. Evidence https://www.census.gov/programs-surveys/acs/. Synthesis No. 142. AHRQ Publication No. 14–05212-EF-1. Rockville, MD: 26. United States Department of Agriculture: Atlas of Rural and Small-Town Agency for Healthcare Research and Quality. In., vol. 14–05212-EF-1; 2016. America. 2018 http://www.ers.usda.gov/data-products/atlas-of-rural-and- 4. Centers for Disease Control and Prevention: Treatment regimens for latent small-town-america.aspx. TB infection (LTBI). 2017. https://www.cdc.gov/tb/topic/treatment/ltbi.htm. 27. Centers for Disease Control and Prevention. Targeted tuberculin testing and 5. Centers for Disease Control and Prevention: Division of Tuberculosis treatment of latent tuberculosis infection. American Thoracic Society. Elimination Strategic Plan 2016–2020. 2015. http://www.cdc.gov/tb/about/ MMWR. 2000;49(RR-6):1–51. strategicplan.htm. 28. Williams R. Generalized ordered logit/partial proportional odds models for 6. US Preventive Services Task Force: Update on Methods: Estimating Certainty and ordinal dependent variables. Stata J. 2006;6(1):58–82. Magnitude of Net Benefit. 2017. http://www.uspreventiveservicestaskforce.org/ 29. Saltelli A, Tarantola S, Campolongo F, Ratto M. Sensitivity analysis in Page/Name/update-on-methods-estimating-certainty-and-magnitude-of-net- practice: A Guide to Assessing Scientific Methods. West Sussex, England: benefit. John Wiley & Sons Ltd; 2004. 7. US Preventive Services Task Force, Bibbins-Domingo K, Grossman DC, Curry 30. IBM Corporation: IBM SPSS modeler 17 algorithms guide. Chicago, IL: IBM SJ, Bauman L, Davidson KW, Epling JWJ, Garcia FA, Herzstein J, Kemper AR, software Group; 2015. et al. Screening for latent tuberculosis infection in adults: US preventive 31. StataCorp: Stata Statistical Software: Release 14. College Station, TX: services task force recommendation statement. JAMA. 2016;316(9):962–9. StataCorp LP; 2016. 8. Sterling TR, Bethel J, Goldberg S, Weinfurter P, Yun L, Horsburgh CR, 32. Horgan CM,Merrick EL,Stewart MT, Scholle SH, Shih S.Improving medication Tuberculosis Epidemiologic Studies C. The scope and impact of treatment management of depression in health plans. Psychiatr Serv. 2008;59(1):72–7. of latent tuberculosis infection in the United States and Canada. Am J 33. Alexander BD, Broccolo BM, Daniels AR, SM DV, Hutto CB, Colborn-Loepere Respir Crit Care Med. 2006;173(8):927–31. C, Matyas DE, Mayo TW, Miles JJ, Polikoff EA, et al. Fundamentals of Health 9. Ehman M, Flood J, Barry PM. Tuberculosis treatment managed by providers Law, vol. 5th. Washington, DC: American Health Laywers Association; 2011. outside the public health department: lessons for the affordable care act. 34. National Committee for Quality Assurance: HEDIS & Performance Measures. PLoS One. 2014;9(10):e110645. 2018. http://www.ncqa.org/hedis-quality-measurement. 10. Institute of Medicine Committee on the Elimination of Tuberculosis in the 35. National Committee for Quality Assurance: Health Plan Accreditation. 2017. United States. Ending neglect: the elimination of tuberculosis in the U.S. http://www.ncqa.org/Programs/Accreditation/Health-Plan-HP.aspx. Washington (DC): national academy of sciences; 2000. 36. Pharmacy Benefits Management Institute: 2015–2016 Prescription Drug 11. Balaban V, Marks SM, Etkind SC, Katz DJ, Higashi J, Flood J, Cronin A, Ho CS, Benefit Cost and Plan Design Report. 2017. https://www.pbmi.com/ Khan A, Chorba T. Tuberculosis elimination efforts in the United States in ItemDetail?iProductCode=2015-2016_BDR&Category=BDR. the era of insurance expansion and the affordable care act. Public Health 37. Brown TA, Rickles NM. Mail-order versus local pharmacies on adherence: Rep. 2015;130(4):349–54. study methods make for unfair comparison. Am J Manag Care. 2010;16(3): 12. The Patient Protection and Affordable Care Act, Sec. 2713, Coverage of 226–7. author reply 227-228 Preventive Services. 2010. https://www.gpo.gov/fdsys/pkg/BILLS- 38. Sandgren AA, Noordegraaf-Schouten MV, van Kessel F, Stuurman A, 111hr3590enr/pdf/BILLS-111hr3590enr.pdf. Oordt-Speets A, van der Werf MJ. Initiation and completion rates for 13. Barnett JCV, Marina S: Health insurance coverage in the United States. 2016 latent tuberculosis infection treatment: a systematic review. BMC http://www.census.gov/content/dam/Census/library/publications/2016/ Infect Dis. 2016;16(1):204. demo/p60-257.pdf. 14. Foutz J, Damico A, Squires E, Garfield R. The Uninsured: A Primer - Key Facts 39. Sharma SK, Sharma A, Kadhiravan T, Tharyan P. Rifamycins (rifampicin, about Health Insurance and the Uninsured Under the Affordable Care Act. rifabutin and rifapentine) compared to isoniazid for preventing tuberculosis The Henry J. Kaiser Family Foundation. 2017.https://www.kff.org/uninsured/ in HIV-negative people at risk of active TB. Cochrane Database Syst Rev. report/the-uninsured-a-primer-key-facts-about-health-insurance-and-the- 2013;7:Cd007545. uninsured-under-the-affordable-care-act/. 40. Lewinsohn DM, Leonard MK, LoBue PA, et al. Official American Thoracic 15. Hirsch-Moverman Y, Daftary A, Franks J, Colson PW. Adherence to treatment for Society/Infectious Diseases Society of America/Centers for Disease Control latent tuberculosis infection: systematic review of studies in the US and Canada. and Prevention clinical practice guidelines: diagnosis of tuberculosis in The international journal of tuberculosis and lung disease. 12(11):1235–54. adults and children. Clin Infect Dis. 2017;64(2):e1-e33. 16. Stuurman AL, Noordegraaf-Schouten MV, Van Kessel F, Oordt-Speets A, 41. Arbuckle C, Tomaszewski D, Aronson BD, Brown L, Schommer J, Morisky Sandgren AA, Van der Werf MJ. Interventions for improving adherence to D, Linstead E. Evaluating factors impacting medication adherence Stockbridge et al. BMC Public Health (2018) 18:662 Page 13 of 13 among rural, urban, and suburban populations. J Rural Health. 2018; 58. Seabury SA, Lakdawalla DN, Dougherty JS, Sullivan J, Goldman DP. https://doi.org/10.1111/jrh.12291. Medication adherence and measures of health plan quality. Am J Manag 42. Francke AL, Smit MC, de Veer AJE, Mistiaen P. Factors influencing the Care. 2015;21(6):e379–89. implementation of clinical guidelines for health care professionals: a systematic 59. Kirkman MS, Rowan-Martin MT, Levin R, Fonseca VA, Schmittdiel JA, Herman meta-review. BMC Medical Informatics and Decision Making. 2008;8(38) WH, Aubert RE. Determinants of adherence to diabetes medications: findings from a large pharmacy claims database. Diabetes Care. 2015;38(4):604–9. 43. McKinlay JB, Link CL, Freund KM, Marceau LD, O’Donnell AB, Lutfey KL. 60. Doshi JA, Lim R, Li P, Young PP, Lawnicki VF, Troxel AB, Volpp KG. Sources of variation in physician adherence with clinical guidelines: results Synchronized prescription refills and medication adherence: a retrospective from a factorial experiment. J Gen Intern Med. 2007;22(3):289–96. claims analysis. Am J Manag Care. 2017;23(2):98–104. 44. Oren E, Winston CA, Pratt R, Robison VA, Narita M. Epidemiology of urban 61. Medicare Payment Advisory Commission: Chapter 7: Measuring the effects tuberculosis in the United States, 2000-2007. Am J Public Health. 2011; of medication adherence for the Medicare population. Report to the 101(7):1256–63. Congress: Medicare and the Health Care Delivery System. 2014. http://www. 45. Gardner T: Urban-suburban migration in the United States, 1955–2000. US medpac.gov/-documents-/reports. Census Bureau Center for Economic Studies Paper No. CES-WP-16-08: 2016. 62. Andrade SE, Kahler KH, Frech F, Chan KA. Methods for evaluation of https://doi.org/10.2139/ssrn.2726211. medication adherence and persistence using automated databases. 46. Centers for Disease Control and Prevention: Official Authorized Addenda: Pharmacoepidemiol Drug Saf. 2006;15(8):565–74. discussion 575-567 Human Immunodeficiency Virus Infection Codes and Official Guidelines for Coding and Reporting ICD-9-CM. 1994. https://wonder.cdc.gov/wonder/ prevguid/m0032908/m0032908.asp. 47. Li JJ. Adherence to treatment of latent tuberculosis infection in a clinical population in new York City. Int J Infect Dis. 2010;14(4):e292–7. 48. Nyamathi AM, Christiani A, Nahid P, Gregerson P, Leake B. A randomized controlled trial of two treatment programs for homeless adults with latent tuberculosis infection. Int J Tuberc Lung Dis. 2006;10(7):775–82. 49. Fiske CT, Yan FX, Hirsch-Moverman Y, Sterling TR, Reichler MR. Risk factors for treatment default in close contacts with latent tuberculous infection. Int J Tuberc Lung Dis. 2014;18(4):421–7. 50. Young KH, Ehman M, Reves R, Peterson BL, Khan A, Chorba TL, Jereb J. Tuberculosis contact investigations–United States, 2003-2012. MMWR Morb Mortal Wkly Rep. 2016;64(50–51):1369–74. 51. McGinnis B, Kauffman Y, Olson KL, Witt DM, Raebel MA. Interventions aimed at improving performance on medication adherence metrics. Int J Clin Pharm. 2014;36(1):20–5. 52. Steiner JF, Prochazka AV. The assessment of refill compliance using pharmacy records: methods, validity, and applications. J Clin Epidemiol. 1997;50(1):105–16. 53. Agency for Healthcare Research and Quality: Measure Summary: Medication management for people with asthma (NQMC: 010514). 2015. https://www.qualitymeasures.ahrq.gov/summaries/summary/49707/ medication-management-for-people-with-asthma-percentage-of- members-5-to-85-years-of-age-during-the-measurement-year-who-were- identified-as-having-persistent-asthma-and-who-were-dispensed-an- asthma-controller-medication-that-they-remained-on-for-at-least-75-of-?q= medication+management+for+people+with+asthma. 54. Agency for Healthcare Research and Quality: Measure Summary: Antidepressant medication management (effective acute phase treatment) (NQMC: 010531). 2015. https://www.qualitymeasures.ahrq.gov/summaries/ summary/49729/antidepressant-medication-management-effective-acute- phase-treatment-percentage-of-members-18-years-of-age-and-older-who- were-treated-with-antidepressant-medication-had-a-diagnosis-of-major- depression-and-who-remained-on-an-antidepressant-medication-for-at- lea?q=antidepressant+medication+management. 55. Agency for Healthcare Research and Quality: Measure Summary: Antidepressant medication management (effective continuation phase treatment) (NQMC: 010532). 2015. https://www.qualitymeasures.ahrq. gov/summaries/summary/49730/antidepressant-medication-management- effective-continuation-phase-treatment-percentage-of-members-18-years- of-age-and-older-who-were-treated-with-antidepressant-medication-had- a-diagnosis-of-major-depression-and-who-remained-on-an-antidepressant- medication-for?q=antidepressant+medication+management. 56. Agency for Healthcare Research and Quality: Measure Summary: Adherence to antipsychotic medications for individuals with schizophrenia (NQMC: 010540). 2015. https://www.qualitymeasures. ahrq.gov/summaries/summary/49738/adherence-to-antipsychotic- medications-for-individuals-with-schizophrenia-percentage-of-members- 19-to-64-years-of-age-during-the-measurement-year-with- schizophrenia-who-were-dispensed-and-remained-on-an-antipsychotic- t-of-their. medication-for-at-least-80-percen 57. Sueta CA, Rodgers JE, Chang PP, Zhou L, Thudium EM, Kucharska-Newton AM, Stearns SC. Medication adherence based on part D claims for patients with heart failure after hospitalization (from the atherosclerosis risk in communities study). Am J Cardiol. 2015;116(3):413–9.

Journal

BMC Public HealthSpringer Journals

Published: May 29, 2018

References

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