TY - JOUR
AU - Gutierrez, Peter M
AB - Abstract Background Veterans with serious mental illnesses face increased risk of both poor medication adherence and death by overdose or suicide. Blister packaging of medications has been associated with population reductions in suicide in the United Kingdom and may improve adherence. Yet, the lack of studies of cost and outcomes has led to controversy about its use. Objective To conduct a cost-utility analysis of blister packaging versus bulk dispensing using in a randomized controlled trial among Veterans with serious mental illnesses or substance abuse. Methods We conducted a cost-utility analysis of a pragmatic randomized controlled trial of 303 Veterans (243 of whom completed the 12-month trial) with post-traumatic stress disorder, schizophrenia, bipolar disorder or major affective disorder in a large VA academic medical center in the West from the perspective of the VA. Quality adjusted life years (QALYs) were estimated from a monthly SF-36. Total healthcare utilization and costs were derived from the VHA Corporate Data Warehouse. Incremental cost-utility ratio and probabilistic sensitivity analysis were estimated. Costs were modeled with regression. Results Mean total VA healthcare costs (including inpatient, outpatient and pharmacy costs) for intervention and control groups were not statistically different at $28 591 and $30 732, respectively. All subjects survived the 12-month trial period, with similar mean QALYs (0.59 and 0.58). Conclusions In this group of veterans with serious mental illnesses, our results suggest that blister packaging all outpatient prescription medications was budget-neutral. Blister packaging may be a cost-effective barrier to intentional overdose with prescribed medications. pharmaco-economics Background and objectives Blister packaging all medications is associated with improved adherence to pharmacotherapy and reductions in suicide, yet the additional cost and lack of evidence of cost-effectiveness have impeded its use.[1–9] In the United Kingdom, a regulatory change requiring a switch from bulk packaging of paracetamol (acetaminophen (Tylenol)) to blister packaging has been associated with reductions in suicide.[1,2] Blister packaging is also associated with improved adherence.[3] and eliminates the need for patients to split tablets for half-dose prescriptions, a known barrier to adherence.[4–6] Most recently, Gutierrez and colleagues conducted a randomized controlled trial in the US Department of Veterans Health Affairs and found that among veterans with post-traumatic stress disorder, bipolar disorder, schizophrenia, substance use disorder or major affective disorder, adherence improved significantly for those with the worst baseline adherence randomized to blister packaging versus dispensing as usual in bulk vials.[7,8] Yet, studies of the cost–utility of blister packaging are lacking and the use of monitored dosage systems (including blister packaging) is controversial due to the marginal cost associated with repackaging and lack of data about cost-savings from avoided hospitalizations and overall cost-effectiveness.[9] The purpose of this study was to perform a cost–utility analysis of blister packaging versus dispensing as usual (DAU) (bulk dispensing in vials) using data from the clinical trial among veterans with serious mental illnesses conducted by Gutierrez and colleagues.[7,8] Methods Study design Design Cost–utility analysis of a randomized controlled trial. Target population and subgroups We conducted a cost–utility analysis of a pragmatic randomized controlled trial in which 303 veterans with schizophrenia, bipolar disorder, major affective disorder and post-traumatic stress disorder were randomized to all prescriptions in blister packaging or in bulk vials (dispensing as usual (DAU)). The trial was designed for these participants because veterans and patients with these disorders are at elevated risk for both poor adherence and suicide.[7,8] The trial and clinical outcomes have been previously described in detail.[7,8] Briefly, participants were 303 veterans aged 18 or older discharged from a psychiatric inpatient unit or receiving services in the outpatient mental health, substance abuse or post-traumatic stress syndrome (PTSD) clinics of a large, urban VA medical centre in the Western United States between September 2011 and January 2014. Inclusion criteria included a diagnosis of one or more of the following: major affective disorder, bipolar affective disorder, post-traumatic stress disorder or schizophrenia. Exclusion criteria included enrolment in any other intervention study, active duty in any military branch, being unable to independently manage medications, non-English speaking or incorrectly answering questions confirming understanding on the consent form. After giving informed consent, participants returned with all of their medications for a baseline assessment, randomization and repackaging of all medications. The pharmacy staff randomized participants to intervention or control using a schedule created in advance with a random number generator. Of the 303 enrolled patients, 243 (80.2%) completed the 12-month trial with at least two assessments.[7] There were no significant differences between those who dropped out and those who completed follow-up. We excluded a participant who experienced several hospitalizations for a total of 186 days during and immediately after which he was unable to participate. At randomization, there were no differences between trial arms by age, diagnoses, race, education and years of service or rank.[7] The typical participant was aged 55, white, unmarried, with some college education and had served in the army for 5 years. Setting and location The Department of Veterans Health Affairs (VHA) is the largest health system in the United States serving more than 6 million veterans through government owned and operated facilities that provide comprehensive healthcare services. Study perspective The perspective is that of the VHA. Comparators Unit dose blister packaging was compared with dispensing as usual (i.e. bulk dispensing in vials) of all prescription medications regardless of indication. Cold-sealed blister cards with 28, 31 or 90 blisters were labelled with the day, date and time of each dose and one dose per blister. Medications prescribed ‘take as needed’ (TPN) were each packaged on a single card. For prescriptions of half-tablet doses, the pharmacy staff used the same manual tablet splitter routinely provided to patients to split tablets before blister packaging. Time horizon We chose 12 months as our time horizon because we would expect to see significant differences in quality of life within 12 months whether blister packaging improved adherence. Discount rate Discounting was not used because costs and outcomes were not measured beyond 12 months. Choice of outcome Quality-adjusted life years (QALYs) were selected as the outcome measure because blister packaging every prescribed medication for middle-aged veterans with serious mental illnesses may improve a variety of outcomes.[11] In addition, the willingness to pay for an additional QALY is known (i.e. $50 000 to $300 000/QALY in the United States).[12,13] Measurement of effectiveness The use of data from only one trial is sufficient for this analysis because the data are drawn from a randomized controlled trail of hundreds of patients in the largest health system in the United States, and to the best of our knowledge, it is the only trial data available about this topic. Estimation of QALYs To calculate the QALY, we first estimated utilities for each patient at each follow-up using the SF-6D, a health state classification system based on 6 dimensions of the SF-36.[14] Participants completed the SF-36 v. 2[15] at baseline and then monthly for 12 months. We calculated a summary preference score for each SF-36 survey.[16] Cumulative quality-adjusted life years (QALYs) over the period of the trial were calculated as the area under the preference score curve plotted against time.[12] The minimally important difference (MID) (i.e. the smallest difference important to patients and clinicians) for the SF-6D is 0.033.[17] The alternative economic definition of MID (any difference that we are willing to pay to modify) specifies a value as small as 0.005 if it is associated with a cost of treatment of 1.0 as defined by the study data.[12] Intervention materials units and unit costs Intervention costs included the costs of the blister packaging materials and the labour to fill the blister packages. Memory Pac® cold-seal blister packages were acquired from Healthcare Logistics, Inc. (http://shop.gohcl.com). Three types were used, based on the prescribed doses per day (Table 1). For 30-day prescriptions of one dose per day, a 31-blister card was used and 1 card dispensed per month. For prescriptions with up to 4 doses per day, a 28-blister card was used per week or 4 cards per 30-day prescription. Vitamins and supplements are routinely prescribed in 90-day supplies, and each of these was packaged in a 90-blister card. Table 1 Units and unit costs (US$) per dispensed prescription . Unit cost ($, 2016) . Source . Blister packaging, per prescription Filling blister card, labour only (per 31 count card) 1.05 X bulk dispensing cost Expert opinion Memory Pac® 31 Count Blister Card $0.46 VA RCT purchasing records, 2012 Memory Pac® 28 Count Weekly Blister Card $0.48 VA RCT purchasing records, 2012 Memory Pac® 90 Count Blister Card $0.52 VA RCT purchasing records, 2012 Bulk dispensing, per prescription Dispensing, labour only (per 30-day fill) 7.92 Decision Support System Bottle and cap (Friendly and Safe Vial with Child Resistant Cap Attached, 16 Dram) $0.02 http://www.gohcl.com/ Accessed 1.5.17 . Unit cost ($, 2016) . Source . Blister packaging, per prescription Filling blister card, labour only (per 31 count card) 1.05 X bulk dispensing cost Expert opinion Memory Pac® 31 Count Blister Card $0.46 VA RCT purchasing records, 2012 Memory Pac® 28 Count Weekly Blister Card $0.48 VA RCT purchasing records, 2012 Memory Pac® 90 Count Blister Card $0.52 VA RCT purchasing records, 2012 Bulk dispensing, per prescription Dispensing, labour only (per 30-day fill) 7.92 Decision Support System Bottle and cap (Friendly and Safe Vial with Child Resistant Cap Attached, 16 Dram) $0.02 http://www.gohcl.com/ Accessed 1.5.17 Open in new tab Table 1 Units and unit costs (US$) per dispensed prescription . Unit cost ($, 2016) . Source . Blister packaging, per prescription Filling blister card, labour only (per 31 count card) 1.05 X bulk dispensing cost Expert opinion Memory Pac® 31 Count Blister Card $0.46 VA RCT purchasing records, 2012 Memory Pac® 28 Count Weekly Blister Card $0.48 VA RCT purchasing records, 2012 Memory Pac® 90 Count Blister Card $0.52 VA RCT purchasing records, 2012 Bulk dispensing, per prescription Dispensing, labour only (per 30-day fill) 7.92 Decision Support System Bottle and cap (Friendly and Safe Vial with Child Resistant Cap Attached, 16 Dram) $0.02 http://www.gohcl.com/ Accessed 1.5.17 . Unit cost ($, 2016) . Source . Blister packaging, per prescription Filling blister card, labour only (per 31 count card) 1.05 X bulk dispensing cost Expert opinion Memory Pac® 31 Count Blister Card $0.46 VA RCT purchasing records, 2012 Memory Pac® 28 Count Weekly Blister Card $0.48 VA RCT purchasing records, 2012 Memory Pac® 90 Count Blister Card $0.52 VA RCT purchasing records, 2012 Bulk dispensing, per prescription Dispensing, labour only (per 30-day fill) 7.92 Decision Support System Bottle and cap (Friendly and Safe Vial with Child Resistant Cap Attached, 16 Dram) $0.02 http://www.gohcl.com/ Accessed 1.5.17 Open in new tab To estimate the incremental labour cost for filling blister packages, we followed VHA guidelines for economic evaluations.[10] The VHA Corporate Data Warehouse (CDW) Decision Support System (DSS) is the national cost-accounting system that approximates opportunity costs. The DSS pharmacy data set lists separate costs for ingredients and dispensing. We estimated the cost of dispensing a blister pack at 5% more than DAU and performed sensitivity analysis for a range of 0–10%. These estimates were provided by the pharmacy manager who supervised the study dispensing. Use of expert opinion is consistent with VA recommendations for economic analyses of healthcare interventions.[18] Healthcare costs We included all VA healthcare costs using the VA Corporate Data Warehouse Decision Support System (DSS) (Table 2): prescription medications (ingredients and dispensing costs) and inpatient and outpatient costs. We excluded costs related to study administration. Table 2 Utilization and cost (US$) of VA health services during the trial, by treatment Group . Blister group, mean (SD), (n = 119) . DAU, mean (SD), (n = 123) . Utilization of VA health care Inpatient Admissions (mean, SD) 0.34 (0.85) 0.36 (0.77) Length of stay, days (mean, SD) 7.00 (22.40) 11.00 (40.20) Pharmacy Total fills (mean, SD) 32.9 (24.2) 36.5 (29.7) 30-day fills 18.8 (16.1) 20.5 (16.4) 90-day fills 7.8 (8.1) 8.9 (8.6) Utilization of VA health care Total inpatient 5946 (17 324) 6738 (17 337) Total outpatient 21 173 (16 651) 22 402 (20 921) Behavioural health 9710 (9205) 11 194 (12 163) Emergency department 524 (760) 602 (1068) Total pharmacy Ingredient cost 1473 (1703) 1592 (2390) Dispensing cost (labour, packaging) 275 (202) 316 (249) . Blister group, mean (SD), (n = 119) . DAU, mean (SD), (n = 123) . Utilization of VA health care Inpatient Admissions (mean, SD) 0.34 (0.85) 0.36 (0.77) Length of stay, days (mean, SD) 7.00 (22.40) 11.00 (40.20) Pharmacy Total fills (mean, SD) 32.9 (24.2) 36.5 (29.7) 30-day fills 18.8 (16.1) 20.5 (16.4) 90-day fills 7.8 (8.1) 8.9 (8.6) Utilization of VA health care Total inpatient 5946 (17 324) 6738 (17 337) Total outpatient 21 173 (16 651) 22 402 (20 921) Behavioural health 9710 (9205) 11 194 (12 163) Emergency department 524 (760) 602 (1068) Total pharmacy Ingredient cost 1473 (1703) 1592 (2390) Dispensing cost (labour, packaging) 275 (202) 316 (249) * P < 0.05 Open in new tab Table 2 Utilization and cost (US$) of VA health services during the trial, by treatment Group . Blister group, mean (SD), (n = 119) . DAU, mean (SD), (n = 123) . Utilization of VA health care Inpatient Admissions (mean, SD) 0.34 (0.85) 0.36 (0.77) Length of stay, days (mean, SD) 7.00 (22.40) 11.00 (40.20) Pharmacy Total fills (mean, SD) 32.9 (24.2) 36.5 (29.7) 30-day fills 18.8 (16.1) 20.5 (16.4) 90-day fills 7.8 (8.1) 8.9 (8.6) Utilization of VA health care Total inpatient 5946 (17 324) 6738 (17 337) Total outpatient 21 173 (16 651) 22 402 (20 921) Behavioural health 9710 (9205) 11 194 (12 163) Emergency department 524 (760) 602 (1068) Total pharmacy Ingredient cost 1473 (1703) 1592 (2390) Dispensing cost (labour, packaging) 275 (202) 316 (249) . Blister group, mean (SD), (n = 119) . DAU, mean (SD), (n = 123) . Utilization of VA health care Inpatient Admissions (mean, SD) 0.34 (0.85) 0.36 (0.77) Length of stay, days (mean, SD) 7.00 (22.40) 11.00 (40.20) Pharmacy Total fills (mean, SD) 32.9 (24.2) 36.5 (29.7) 30-day fills 18.8 (16.1) 20.5 (16.4) 90-day fills 7.8 (8.1) 8.9 (8.6) Utilization of VA health care Total inpatient 5946 (17 324) 6738 (17 337) Total outpatient 21 173 (16 651) 22 402 (20 921) Behavioural health 9710 (9205) 11 194 (12 163) Emergency department 524 (760) 602 (1068) Total pharmacy Ingredient cost 1473 (1703) 1592 (2390) Dispensing cost (labour, packaging) 275 (202) 316 (249) * P < 0.05 Open in new tab Currency, price date and conversion All costs were standardized and reported in 2016 US dollars using the Consumer Price Index medical care values as follows: 3.7% (over 2012), 2.5% (over 2013), 2.4% (over 2014) and 2.6% (over 2015).[19] Analysis Costs and utilization Two-tailed t-tests were used to compare mean utilization and costs by trial arm across a variety of healthcare services. A generalized linear model with a gamma family and log link was used to characterize total costs by trial arm controlling for the intervention, hospitalization in the previous 12 months, hospitalization length of stay during the trial period, sex, age, education, number of prescriptions at baseline and comorbidities. A modified Parks test confirmed the choice of the gamma family. Incremental cost–utility ratio The incremental cost–utility ratio was calculated as: CostA−CostBQALYSA−QALYSBCostA−CostBQALYSA−QALYSB Acceptability curve and sensitive analyses Bootstrapping was used to create an acceptability curve illustrating the probability that the estimated ratio falls below cost-per-QALY willingness to pay thresholds of $50 000 and $300 000 per QALY respectively. Sensitivity analyses were also included for significant VA costs that are not typically considered healthcare costs in the literature, specifically supported employment and community services to support behavioural health patients. We performed all analyses with these costs and then again without them to identify any influence on total costs and the incremental cost–utility ratio. Hospitalizations interrupted trial participation. To address this issue, we re-ran our analysis twice, once including only participants who were never hospitalized during the trial period and once including all participants except for an influential outlier hospitalized for 168 days. This research was reviewed and approved by the Colorado Multiple Institution Review Board, the Department of Defense Human Research Protection Office and the Department of Veterans Affairs Office of Research and Development and Participant Safety Subcommittee. Results Mortality and quality of life All participants survived the study period. Summary preference scores for each of the 2242 completed SF-36 forms ranged from 0.30 to 1.0 (SD 0.12). Mean cumulative QALYs were similar across the blister packaging and dispensing as usual groups (0.59 (SD 0.08) versus 0.58 (SD 0.08)) (Table 3). The mean incremental QALY approached but did not reach statistical significance (0.11, 95% CI −0.008 – 0.031). Table 3 Utilities and the associated mean cumulative QALY by group for the 12-month trial period . Blister packaging, mean (SD) . Dispensing as usual (DAU), mean (SD) . SF6D-derived utilities Baseline 0.575 (0.112) 0.573 (0.093) Month 3 0.590 (0.107) 0.582 (0.103) Month 6 0.598 (0.109) 0.576 (0.099) Month 9 0.597 (0.094) 0.590 (0.084) Month 12 0.597 (0.051) 0.586 (0.036) Mean cumulative QALYs for the trial follow-up period 0.591 (0.081) 0.580 (0.075) Mean incremental QALYs (95% CI) 0.011 (−0.008 – 0.031) . Blister packaging, mean (SD) . Dispensing as usual (DAU), mean (SD) . SF6D-derived utilities Baseline 0.575 (0.112) 0.573 (0.093) Month 3 0.590 (0.107) 0.582 (0.103) Month 6 0.598 (0.109) 0.576 (0.099) Month 9 0.597 (0.094) 0.590 (0.084) Month 12 0.597 (0.051) 0.586 (0.036) Mean cumulative QALYs for the trial follow-up period 0.591 (0.081) 0.580 (0.075) Mean incremental QALYs (95% CI) 0.011 (−0.008 – 0.031) Open in new tab Table 3 Utilities and the associated mean cumulative QALY by group for the 12-month trial period . Blister packaging, mean (SD) . Dispensing as usual (DAU), mean (SD) . SF6D-derived utilities Baseline 0.575 (0.112) 0.573 (0.093) Month 3 0.590 (0.107) 0.582 (0.103) Month 6 0.598 (0.109) 0.576 (0.099) Month 9 0.597 (0.094) 0.590 (0.084) Month 12 0.597 (0.051) 0.586 (0.036) Mean cumulative QALYs for the trial follow-up period 0.591 (0.081) 0.580 (0.075) Mean incremental QALYs (95% CI) 0.011 (−0.008 – 0.031) . Blister packaging, mean (SD) . Dispensing as usual (DAU), mean (SD) . SF6D-derived utilities Baseline 0.575 (0.112) 0.573 (0.093) Month 3 0.590 (0.107) 0.582 (0.103) Month 6 0.598 (0.109) 0.576 (0.099) Month 9 0.597 (0.094) 0.590 (0.084) Month 12 0.597 (0.051) 0.586 (0.036) Mean cumulative QALYs for the trial follow-up period 0.591 (0.081) 0.580 (0.075) Mean incremental QALYs (95% CI) 0.011 (−0.008 – 0.031) Open in new tab VA healthcare utilization and cost Mean VA healthcare utilization and costs during the trial period were similar for both the blister packaging and DAU groups on all measures based on two-tailed tests with a P-value of <0.05 (Table 2). A generalized linear regression model (GLM) with gamma family and log link function of the total costs per participant by study arm met the modified Parks test for the gamma family and performed relatively well. Model results indicated that the intervention was not a statistically significant explanatory factor for individual variation in total VA healthcare costs. Rather, individual differences were attributable to male sex and older age, total days spent in the hospital in the 12 months prior to the study period and history of drug abuse, bipolar disorder or major affective disorder. Cost–utility analysis and probabilistic sensitivity analysis For the 12-month trial period, the mean total cost in the intervention group was $28 591 and in the DAU group, $30 732 (Table 4). Mean quality-adjusted life years (QALYs) were 0.59 and 0.58 and were not statistically different at P < 0.05. The point estimates suggest that blister packaging was associated with both lower total healthcare costs and better quality of life, but this result may have been due to chance as it was not statistically significant at P < 0.05. At both low and high thresholds of willingness to pay ($50 000/QALY and $300 000/QALY), the probability of blister packaging being cost-effective was similar (Figure 1). That is, there is no value of willingness to pay for which we can be 95% confident that blister packaging is preferred over dispensing as usual. Table 4 Costs, outcomes, cost–utility and the acceptability curve Group . Mean total cost ($ 2016) . Cumulative quality-adjusted life years, Mean . ICER . P(CE) at $50 000 (WTP) . P(CE) at $300 000 (WTP) . Ceiling ratio at p(CE) = 0.95 . Blister pack (n = 119) 28 591 0.59 NA 0.775 0.878 NA DAU (n = 123) 30 732 0.58 NA NA NA NA Difference (95% CI) −2140 (−9053 – 4773) 0.01 (−0.01 – 0.03) Blister packaging approaches dominance but not statistically significant NA NA NA Group . Mean total cost ($ 2016) . Cumulative quality-adjusted life years, Mean . ICER . P(CE) at $50 000 (WTP) . P(CE) at $300 000 (WTP) . Ceiling ratio at p(CE) = 0.95 . Blister pack (n = 119) 28 591 0.59 NA 0.775 0.878 NA DAU (n = 123) 30 732 0.58 NA NA NA NA Difference (95% CI) −2140 (−9053 – 4773) 0.01 (−0.01 – 0.03) Blister packaging approaches dominance but not statistically significant NA NA NA NA, Not applicable. Open in new tab Table 4 Costs, outcomes, cost–utility and the acceptability curve Group . Mean total cost ($ 2016) . Cumulative quality-adjusted life years, Mean . ICER . P(CE) at $50 000 (WTP) . P(CE) at $300 000 (WTP) . Ceiling ratio at p(CE) = 0.95 . Blister pack (n = 119) 28 591 0.59 NA 0.775 0.878 NA DAU (n = 123) 30 732 0.58 NA NA NA NA Difference (95% CI) −2140 (−9053 – 4773) 0.01 (−0.01 – 0.03) Blister packaging approaches dominance but not statistically significant NA NA NA Group . Mean total cost ($ 2016) . Cumulative quality-adjusted life years, Mean . ICER . P(CE) at $50 000 (WTP) . P(CE) at $300 000 (WTP) . Ceiling ratio at p(CE) = 0.95 . Blister pack (n = 119) 28 591 0.59 NA 0.775 0.878 NA DAU (n = 123) 30 732 0.58 NA NA NA NA Difference (95% CI) −2140 (−9053 – 4773) 0.01 (−0.01 – 0.03) Blister packaging approaches dominance but not statistically significant NA NA NA NA, Not applicable. Open in new tab Figure 1 Open in new tabDownload slide Cost-effectiveness acceptability curve at 50 000 and 300 000 per QALY: blister packaging and dispensing as usual demonstrate similar treatment value in this group of veterans with serious mental illness and comorbidities. Excluding VA supported employment and community support services for behavioural health patients resulted in lower total costs, but did not change the cost–utility of the intervention compared with DAU. Subanalyses including only participants who were never hospitalized produced results similar to our primary finding. Discussion To the best of our knowledge, this is the first pragmatic randomized trial to study blister packaging and its effects on QALYs among participants with serious mental illnesses. This secondary analysis of a pragmatic randomized trial of veterans discharged from the psychiatric inpatient unit or receiving services in outpatient mental health, substance abuse or PTSD clinics found that the costs and outcomes of blister packaging and bulk dispensing were similar. Previous analyses of this trial data found that the least adherent patients at baseline became significantly more adherent to their prescribed medications when provided with blister packaging.[7] Our cost–utility analysis found that blister packaging was dominant over dispensing as usual, but the result did not reach statistical significance. Subanalyses of only participants who were never hospitalized produced similar results. As is typical of trials, this study was underpowered to detect suicide, an extremely rare outcome occurring in fewer than 33.4 per 100 000 veterans.[20] Given the equivocal findings, low cost of the intervention, the VA’s high priority focus on suicide prevention and the UK findings of Hawton and colleagues, blister packaging may continue to be considered as a possible means to both improve patient adherence to pharmacotherapy and impede use of prescribed medications for intentional overdose.[1,2,7] Limitations Only medications filled at VA were included in this study. Yet, the high mean number of prescriptions, diagnostic severity and high utilization of VA services across both arms suggests that these participants may rely on VA exclusively. Large-scale implementation of the intervention would likely involve other technologies, economies of scale and mail order which would all reduce the costs of the blister packaging intervention.[21] In the current study, however, the intervention cost was very small compared with total healthcare costs, and so, any reduction in costs associated with these effects are unlikely to change the finding that blister packaging was similar in cost and outcome to bulk dispensing. Prescriptions with half-tablet administration instructions require the pharmacy staff to split tablets during dispensing when blister packaging but not when dispensing as usual in bulk packaging. In psychiatric settings, tablet splitting is a routine method for adjusting doses, improving swallowing and reducing cost.[22] For example, sertraline (Zoloft) (commonly prescribed to treat depression, obsessive–compulsive disorder and panic and anxiety disorders) may be prescribed as a 30-day supply of fifteen 100mg tablets to be split and one-half taken daily. We did not track the number of split-tablet prescriptions but this was unlikely to have affected the results due to the very small cost of splitting relative to the other costs in the study. Little is known about the effectiveness of pill-splitting by patients versus pharmacy staff, and this question was beyond the scope of our study. Complicated regimens or cognitive impairment may increase the risk that patients neglect to split tablets or split the wrong tablets in their regimens. Patients accustomed to splitting medications as prescribed may at their own initiative split other medications, for example to save money by taking half-doses. Yet, splitting enteric-coated, sustained or controlled-release medications could increase the risk of side effects and reduce effectiveness. Despite this fact, US Pharmacopeia guidelines for the drug content of split tablets do not exist. In fact, a recent report by the Canadian Agency for Drugs and Technologies in Health called for research about both the clinical and economic outcomes associated with pill-splitting.[23] Hospitalizations reduced exposures to the intervention. About one-third of study participants were hospitalized during the 12-month trial with mean lengths of stay of 7 and 11 days in the intervention and DAU groups, respectively, with total inpatient days as high as 80. During periods of hospitalization, patients are not administering their own medications. Thus, the effect of the intervention is weakened. Yet, a subanalysis of participants who were never hospitalized also failed to identify any statistically significant differences in costs, outcomes or cost–utility across the blister packaging and dispensing as usual groups. Like all randomized controlled trials, this trial was underpowered to detect suicide and suicide attempt. Conclusions In summary, in this group of veterans seen for post-traumatic stress disorder, bipolar disorder, major affective disorder or schizophrenia in the US Department of Veterans Health Affairs, blister packaging all medications was no different in cost, mortality, quality of life or cost–utility than dispensing as usual in bulk vials over the 12-month study period. Declarations Conflict of Interest The Author(s) declare(s) that they have no conflicts of interest to disclose. Funding This work was funded, in part, by the Military Suicide Research Consortium (MSRC), an effort supported by the Office of the Assistant Secretary of Defense for Health Affairs under Award No. (W81XWH-10-2-0181). Opinions, interpretations, conclusions and recommendations are those of the authors and are not necessarily endorsed by the CoE, MIRECC, MSRC, the Department of Defense, the Department of Veterans Affairs or the United States Government. Authors’ contributions Jill Lavigne designed the study, directed the analyses, drafted the manuscript and finalized the report. Kristin Falbo performed a literature review and contributed to the introduction. Peter Gutierrez conducted the original trial on which this analysis is based, participated in the design, analysis and interpretation of results and participated in the drafting of the manuscript. Acknowledgements This material is the result of work supported with staff, resources and facilities at the US Department of Veterans Health Affairs Center of Excellence for Suicide Prevention (CoE) and the Military Suicide Research Consortium via the Denver Mental Illness Research Education and Clinical Center (MIRECC). Cathy Kane, MS provided some programming in support of this project. References Hawton K et al. . Long term effect of reduced pack sizes of paracetamol on poisoning deaths and liver transplant activity in England and Wales: interrupted time series analyses . BMJ 2013 ; 346 : f403 . Google Scholar Crossref Search ADS PubMed WorldCat Hawton K et al. . Effects of legislation restricting pack sizes of paracetamol and salicylate on self poisoning in the United Kingdom: before and after study . BMJ 2001 ; 322 : 1203 – 1207 . Google Scholar Crossref Search ADS PubMed WorldCat Conn VS et al. . Packaging interventions to increase medication adherence: systematic review and meta-analysis . Curr Med Res Opin 2015 ; 31 : 145 – 160 . Google Scholar Crossref Search ADS PubMed WorldCat Stone VE et al. . Antiretroviral regimen complexity, self-reported adherence, and HIV patients’ understanding of their regimens: survey of women in the her study . J Acquir Immune Defic Syndr 2001 ; 28 : 124 – 131 . Google Scholar Crossref Search ADS PubMed WorldCat George J et al. . Development and validation of the medication regimen complexity index . Ann Pharmacother 2004 ; 38 : 1369 – 1376 . Google Scholar Crossref Search ADS PubMed WorldCat Martin S et al. . The Antiretroviral Regimen Complexity Index: a novel method of quantifying regimen complexity . J Acquir Immune Defic Syndr 2007 ; 45 : 535 – 544 . Google Scholar Crossref Search ADS PubMed WorldCat Gutierrez PM et al. . Blister packaging medication increases treatment adherence for psychiatric patients . J Psychiatr Pract 2017 ; 23 : 320 – 327 . Google Scholar Crossref Search ADS PubMed WorldCat Gutierrez PM et al. . Blister packaging medication to increase treatment adherence and clinical response: impact on suicide-related morbidity and mortality . In: Lavigne JE, ed. Frontiers in Suicide Risk: Research, Treatment and Prevention . Hauppauge, NY : Nova Science Publishers Inc , 2012 : 195 – 202 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Ubido J , Scott-Samuel A. An evaluation of the cost effectiveness of monitored dosage systems (MDS) as an aid to maintaining independence in taking medication . LHO Report Series #98. October 2014. 23 pages. Published by the Liverpool Public Health Observatory . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Health Services Research and Development Service . QUERI Economic Analysis Guidelines . June, 2014 . http://www.herc.research.va.gov/files/MPDF_773.pdf. Accessed 28 September 2015. Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Zullig LL et al. . A protocol to evaluate the efficacy, perceptions, and cost of a cholesterol packaging approach to improve medication adherence . Contemp Clin Trials 2014 ; 39 : 106 – 112 . Google Scholar Crossref Search ADS PubMed WorldCat Glick HA et al. . Handbooks in Health Economic Evaluation Series: Economic Evaluation in Clinical Trials , 2nd edn. Oxford, New York : Oxford University Press , 2014 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Neumann PJ et al. . Updating cost-effectiveness: the curious resilience of the $50,000-per-QALY threshold . N Engl J Med 2014 ; 371 : 796 – 797 . Google Scholar Crossref Search ADS PubMed WorldCat Brazier JE et al. . Deriving a preference-based single index from the UK SF-36 Health Survey . J Clin Epidemiol 1998 ; 51 : 1115 – 1128 . Google Scholar Crossref Search ADS PubMed WorldCat Ware JE Jr, Sherbourne CD. The MOS 36-item short-form health survey (SF-36).I. Conceptual framework and item selection . Med Care 1992 ; 30 : 473 – 483 . Google Scholar Crossref Search ADS PubMed WorldCat Brazier JE , Roberts J. The estimation of a preference-based measure of health from the SF-12 . Med Care. 2004 ; 42 : 851 – 859 . Google Scholar Crossref Search ADS PubMed WorldCat Walters SJ , Brazier JE. What is the relationship between the minimally important difference and health state utility values? The case of the SF-6D . Health Qual Life Outcomes. 2003 ; 1 : 4 . Google Scholar Crossref Search ADS PubMed WorldCat Smith MW et al. . Microcost methods for determining VA healthcare costs . Menlo Park, CA : Health Economics Resource Center , 2010 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC US Bureau of Labor Statistics . Consumer Price Index: All Urban Consumers, Seasonal, Medical Care . http://data.bls.gov/cgi-bin/surveymost. Accessed 22 October 2015. Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Kang HK et al. . Suicide risk among 1.3 million veterans who were on active duty during the Iraq and Afghanistan wars . Ann Epidemiol 2015 ; 25 : 96 – 100 . Google Scholar Crossref Search ADS PubMed WorldCat McKesson Corporation . Unit Dose vs. Bulk Oral Solid Medication Purchasing Patterns and Repackaging: Sampling and Analysis . Cranberry Township, PA : McKesson Corporation , 2010 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Helmy SA . Tablet splitting: is it worthwhile? Analysis of drug content and weight uniformity for half tablets of 16 commonly used medication in the outpatient setting . J Manag Care Spec Pharm 2015 ; 21 : 76 – 86 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Canadian Agency for Drugs and Technologies in Health . Pill Splitting: A Review of Clinical Effectiveness, Cost-Effectiveness and Guidelines – June 2015 . https://cadth.ca/pill-splitting-review-clinical-effectiveness-cost-effectiveness-and-guidelines. Accessed 15 August 2017. Google Scholar PubMed OpenURL Placeholder Text Google Preview WorldCat COPAC Footnotes † Previous Peer-Reviewed Presentations: A version of this manuscript was presented as a poster at the Military Suicide Research Consortium Annual Meeting in Denver, Colorado in July 2016. © 2019. This article is a U.S. Government work and is in the public domain in the USA This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) © 2019. This article is a U.S. Government work and is in the public domain in the USA
TI - Cost–utility analysis of blister packaging all outpatient medications for veterans with bipolar disorder, major affective disorder, post-traumatic stress disorder or schizophrenia
JF - Journal of Pharmaceutical Health Services Research
DO - 10.1111/jphs.12324
DA - 2019-11-03
UR - https://www.deepdyve.com/lp/oxford-university-press/cost-utility-analysis-of-blister-packaging-all-outpatient-medications-LMbJVfARJd
SP - 401
EP - 406
VL - 10
IS - 4
DP - DeepDyve
ER -