New Start Versus Continuing Users on Aripiprazole: Implications for Policy

, , , ,
The American Journal of Managed Care, January 2015, Volume 21, Issue 1

New starts on aripiprazole were less adherent than continuing users. Prescriptions for 90-day supplies should be reserved for patients who have established effectiveness and tolerance.



To evaluate utilization of 90-day-supply prescriptions of aripiprazole.

Study Design

One year (April 1, 2011, to March 31, 2012) retrospective cohort study from the Veterans Affairs San Diego Healthcare System.


The primary outcome was to determine the difference in adherence for new starts versus continuing users on aripiprazole, as determined by medication possession ratio (MPR). Secondary outcomes included odds of adherence and refilling at least once associated with being a new start. Adherence was defined as MPR ≥0.8. Separate regression models (linear and logistic) were run for the entire population, as well as a subgroup analysis of 90-day prescription patients only.


A total of 749 patients, 328 of whom were new starts, were included in the analysis. Both new starts (41.2%) and continuing users (69.1%) had a large portion who received 90-day supplies. New-start patients had significantly lower MPR than continuing users (—0.13; 95% CI, –0.18 to –0.08). Logistic regressions showed that new starts also had lower odds of adherence (odds ratio [OR], 0.46; 95% CI, 0.33-0.65) and of refilling at least once (OR, 0.43; 95% CI, 0.28-0.66) compared with continuing users.


Patients who were continuing users of aripiprazole were more likely to be adherent and refill their medication. Overutilization of 90-day supplies of high-cost agents, particularly in new starts, may lead to waste. It is recommended that patients newly started on high-cost agents should initially be provided a 30-day-supply prescription until it is established that effectiveness and tolerance have been achieved.

Am J Manag Care. 2015;21(1):e43-e50

Dispensing high-cost agents to new starts as a 90-day supply can lead to excessive waste.

  • Few data are currently available on this topic in the literature.
  • This analysis demonstrates that new starts on aripiprazole are less likely to be adherent or to refill their prescription in comparison with continuing users.
  • Healthcare systems and pharmacy benefits administrators should consider limiting high-cost agents to a 30-day supply or less upon initiation until tolerance and effectiveness can be established in each patient.

The Veterans Health Administration (VA) promotes dispensing 90-day-supply prescriptions for chronic medications via a centralized mail order pharmacy.1 Studies that directly evaluated the effect of prescription size on adherence suggest that an increase in the amount of medication dispensed could positively influence medication adherence.2,3 Mail order pharmacies can also reduce total pharmacy costs by reducing dispensing costs through optimized efficiency and quality practices.1,4 However, dispensing high-cost agents as a 90-day supply runs the risk of high levels of wastage, particularly in patient populations not stabilized on therapy, such as new starts.

There are no formal studies on the amount of drugs wasted, but drug amnesty schemes, in which patients are invited to dispose of unused medicines at their local pharmacy, have demonstrated that up to 20% of medicines returned are unopened.5 Furthermore, studies have suggested that if all returned medicines were limited to a 28-day supply, then wastage would be reduced by one-third.6 Additionally, the amount of a drug wasted also depends on how frequently physicians make changes to patients’ drug therapy and the quantity of a drug patients have on hand when the changes are made. It is clear that dispensing smaller quantities, such as a 30-day supply, can reduce the quantity of medication wasted if a switch in drug therapy is made.7

Federal supply schedule pricing of aripiprazole (Abilify) ranges from $10 to $15 per tablet, resulting in an average cost greater than $1000 per 90-day supply.8 A recent internal review at the Veterans Affairs San Diego Healthcare System (VASDHS) reported that approximately 60% of aripiprazole prescriptions were dispensed as a 90-day supply and accounted for 90% of tablets used.

The purpose of this study was to analyze VASDHS patients receiving aripiprazole to assess medication adherence and refill history by comparing new starts versus continuing users. These measures were thought to serve as a proxy for wastage. We hypothesized that patients newly started on aripiprazole were less likely to be adherent; if so, this could indicate that utilizing a 90-day supply for new starts may be wasteful for high-cost medications such as aripiprazole.


Study Design and Patient Population

The VASDHS provides healthcare to approximately 232,000 veterans in San Diego and Imperial Valley counties.9 This was a 1-year retrospective cohort study of veterans from the VASDHS. Patient data and patient refill history were retrieved from the VA computerized medical records database. Inclusion criteria for this study were limited to veterans with benefits eligibility at VASDHS with a prescription for aripiprazole within the study period (April 1, 2011, to March 31, 2012). Each unique patient was followed for 1 year post index date (defined as the date of first prescription for aripiprazole for each respective patient in the study's time frame). No patients were excluded from the analysis. New start was defined as no aripiprazole in the year prior to the index date. Patients were required to have VA eligibility for the entire 24-month time frame for study inclusion, and were considered to be in the 90-day-supply group based upon whether their initial prescription on the index date was for a 90-day supply.

Pharmacy Benefits Description

The VA has a flat medication co-payment scheme in which a veteran pays no more than $9 for each 30-day supply of medication10; for example, a 90-day supply prescription for aripiprazole would cost no more than $27 for the patient. However, some patients are exempt from co-payments for some or all of their medications. Whether or not a veteran has a co-payment for a medication is dependent upon whether the condition being treated is related to their service time, as well as to the veteran’s total disability due to service (referred to as service connection status).11 If the medication is for treatment of a disability related to their service and has been recognized as such by being deemed a “service-connected condition,” the veteran would have no co-payment for such medications. For example, if a veteran develops posttraumatic stress disorder (PTSD) from their service, they would pay no medication co-payments for any prescriptions related to their PTSD. In addition, if the veteran has lost at least 50% of their earning potential due to their service (ie, being service-connected ≥50%), then the veteran is exempt from co-payments for all medications.

Outcomes and Independent Variables

The primary outcome of this study was to determine the difference in adherence for new start versus continuing users on aripiprazole, as determined by medication possession ratio (MPR). The MPR was calculated as the ratio of days the medication was supplied in the 1-year post index date divided by days in the time interval (in this study, a denominator of 365 was utilized). Secondary outcomes included odds of adherence and refilling at least once. Adherence was defined as MPR ≥0.8. MPR values above 1.0 were truncated to 1.0 in order to prevent overestimation of adherence levels. The independent variables controlled for in the regression models were: age, race, ethnicity, marital status, period of service, mental health disorders, outside insurance, co-payment status, combat veteran status, and day supply (for allpatient groups only, not the 90-day-supply-only subgroup models).

Statistical Methods

Baseline characteristics, as well as an unadjusted analysis of outcomes, were compared between groups using χ2 tests for categorical data and Student's t tests for continuous data. Linear regression was used to measure the primary outcome of change in MPR for new start versus continuing users. For secondary outcome measures when the dependent variable was divided into a dichotomous outcome (adherent or not adherent; refilled or not refilled), logistic regression was utilized to analyze the association between new start versus continuing users. Separate regression models were run for the entire population as well as for a subgroup of 90-day-prescription patients only. Odds ratios (ORs) with 95% CIs and P values were reported. All data analysis was performed using IBM SPSS Statistics, version 18.0 (SPSS Inc, Chicago, Illinois).

RESULTSStudy Population

Table 1

A total of 749 patients met inclusion criteria. At baseline, differences were observed between new starts and continuing users (). New-start aripiprazole patients were younger (P <.001), less likely to have outside insurance (P = .001), less likely to have schizophrenia or depression (P <.001 for both), and more likely to have PTSD (P = .009). The overall makeup of race, ethnicity, marital status, co-payment status, and period of service was not statistically different between groups, although it should be noted that the new starts had more Persian Gulf War veterans (47% vs 40.1%) and fewer Vietnam Era veterans (29.3% vs 37.3%) compared with continuing users.

Unadjusted Results

Table 2

Continuing users were significantly more likely to receive a 90-day prescription compared with new starts (P <.001). However, new starts did still have a large population (41.2%) who received 90-day supplies (Table 1). Continuing users had a higher rate of adherence over new starts, as measured by average MPR (P <.001), as well as number of patients adherent (P <.001). This trend was observed for both the entire population, as well as for the 90-day-supply subgroup. New starts were significantly more likely to be poorly adherent (P <.001), as well as to never refill their prescription (P <.001). This trend was observed for both the entire population, as well as the 90-day-supply subgroup. Unadjusted results are summarized in .

Regression Models Results

Table 3

Overall, patients had significantly lower MPR associated with being a new start (—0.13; 95% CI, –0.18 to –0.08) in the linear regression. Logistic regressions for the entire patient population showed that new starts had lower odds of adherence (OR, 0.46; 95% CI, 0.33-0.65) and for refilling a prescription at least once (OR, 0.43; 95% CI, 0.28-0.66) when compared with continuing users. The subgroup analysis of 90-day-prescription patients showed similar trends and results for all measures ().


Patients who were new starts on aripiprazole were less likely to remain adherent and continue on their medication relative to patients who were continuing users. Although continuing users had a higher proportion of their prescriptions for 90-day supply relative to patients categorized as new starts (69% vs 41%), the practice was fairly common for both groups. These results have important implications in the 90-day-supply fill policy at VASDHS.

Some studies have reported that a 90-day-supply policy may reduce overall operation costs despite a high degree of drug wastage. Domino et al reported that a decrease in the supply from 100-day to 34-day supply resulted in an increase in dispensing fees regardless of a decrease in diverted wastage in the Medicaid population from 1999 to 2000.12 Walton et al reported that the average prescription costs for 30- and 90-day supplies of hydroxymethyl glutaryl coenzyme A (HMG-CoA) reductase-inhibitors were $0.56 and $0.54, respectively.7 The quantity of wastage was reported for the 30- and 90-day supplies as 1.06 days and 5.33 days, respectively. This translated into total unnecessary costs of $5.62 and $3.17, respectively.

However, drug costs have grown substantially since the time frame in which those analyses were conducted. By adopting a blanket policy of 90-day drug dispensing, healthcare systems may potentially invest more on operations and wastage that would exceed potential benefits. Therefore, a 30-day drug dispensing policy may be worth pursuing for specific high-cost medications. Parikh et al reported that a 30-day supply would be more cost-effective when the daily cost of a prescription exceeds $1.50 per day.13 Prescription costs below this threshold would benefit from a 90-day drug dispensing policy through the reduction of staff and mailing costs; however, this benefit would be eliminated with a prescription greater than $1.50 per day. Aripiprazole is approximately $10 to $15 per day for federal facilities with a potential cost saving to the healthcare system if a 30-day drug dispensing policy were in place.

Adopting a 30-day policy would potentially reduce the overall costs to the healthcare system; currently, such decisions are typically made on a drug-by-drug basis. In order to help standardize guidelines for cost-effective medication therapy, one solution may be to define a firm threshold for what is considered a high-cost agent and restrict such medications to 30-day maximum supplies. No such threshold has been defined nationally for the VA healthcare system, with different sites or regions possibly having varying policies or restrictions in this regard.

A national policy or 30-day maximum medication list would help to unify and standardize drug dispensing practices across the country. If such a practice was to be adopted, it would also be wise to define situations for exceptional approvals. Whether different disease states or specialties (eg, oncology, cardiology, psychiatry) “deserve” different supply restrictions is not a well-defined issue and would also need to be included in the discussion of any such policy formation.

Adherence is affected by the medication day supply.2,3 Proponents of the 90-day drug dispensing policy argue that adherence is improved with a 90-day supply, which is also necessary for maintaining or controlling chronic diseases. However, high levels of medication wastage may attenuate this benefit to the healthcare system if the patient is intolerant to the drug therapy, which results in increased wastage. This is especially important with high-cost medications such as aripiprazole, which can cost $900 to $1350 per 90-day supply.

The relationship between adherence and increasing the day supply of aripiprazole will need to be examined in order to determine the benefits of using a 90-day drug dispensing policy. From the perspective of a healthcare system concerned with increasing prescription costs, reducing drug wastage of high-cost agents is a critical goal to an already budget-constrained system. Determining which medications should be allowed a 90-day supply will require local pharmacy and therapeutics committees to analyze the projected drug wastage, operational costs, and clinical implications for their respective institutions or health systems.

Alternatively, having an initial fill policy of 30 days to demonstrate tolerance may identify patients who are less likely to be nonadherent and wasteful with their medication. Several other studies have noted the utilization of 14-day supplies for new starts of high-cost agents that are known for high rates of adverse events. One VA study reported dispensing an initial 14-day supply of dabigatran (Pradaxa) to ensure tolerance prior to dispensing larger quantities.14 Other healthcare systems and pharmacy benefits management companies have reported taking similar initial 14-day supply approaches with oral oncology agents, reporting both cost savings and increases in adherence.15,16 In all cases, the smaller supply was given in conjunction with a more intensive followup process to ensure tolerance and adherence.


There were several limitations to our analysis. First, this study was performed on the veteran population, which may limit generalizability to a nonveteran population. The VA is an integrated system that utilizes mail order pharmacies that may offer advantages in terms of reduced dispensing and staff costs compared with individual community pharmacies; therefore, the costs associated with dispensing and staffing may not reflect those of the community. In addition, it is uncertain how adherence in a veteran population compares with the nonveteran population, which may or may not affect the external validity of our results.

Second, our study used MPR as a measure of patient adherence, which is an indirect measure of patient consumption of medication. True adherence would require direct observation of a patient consuming their medication, which is unlikely to be performed in a general setting. However, McHorney et al reported that MPR was a good proxy for patient adherence, which reduces any potential issues associated with internal validity.17

Finally, our study was an observational study, which was limited to measuring variables that were available in the chart. We were unable to determine whether income status or socioeconomic status would have any potential confounding on the adherence of these patients. As a result, our conclusions should be taken with the caveat that these variables were not controlled for.


Our study reported that patients who were new starts on aripiprazole were less likely to be adherent and to refill their medication. Demonstration of tolerance to aripiprazole may be necessary prior to dispensing a 90-day supply to ensure cost-effective care from a system perspective. It is recommended that patients newly started on aripiprazole or other similar high-cost agents should initially be provided a 30-day-supply prescription to determine if efficacy and safety are achieved. Once this has been performed, then consideration of a larger day supply may be given on a case-by-case basis.Author Affiliations: Pharmacoeconomics & Formulary Management, Veterans Affairs San Diego Healthcare System (RK, JBN, MB, NM, MHT), San Diego, CA.

Source of Funding: None.

Author Disclosures: The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (RK, JBN, MB, NM); acquisition of data (RK, JBN, MB); analysis and interpretation of data (RK, JBN, MHT); drafting of the manuscript (RK, JBN, MB, MHT, NM); critical revision of the manuscript for important intellectual content (RK, JBN); statistical analysis (RK, JBN); administrative, technical, or logistic support (MB, NM, MHT); and supervision (RK).

Address correspondence to: Rashid Kazerooni, PharmD, BCPS, Pharmacoeconomics Program Manager, Veterans Affairs San Diego Healthcare System, 3350 La Jolla Village Dr (119), San Diego, CA 92161. E-mail:

1. Aragon BR, Pierce RA 2nd, Jones WN. VA CMOPs: producing a pattern of quality and efficiency in government. J Am Pharm Assoc (2003). 2012;52(6):810-815.

2. Batal HA, Krantz MJ, Dale RA, Mehler PS, Steiner JF. Impact of prescription size on statin adherence and cholesterol levels. BMC Health Serv Res. 2007;7:175.

3. Steiner JF, Robbins LJ, Roth SC, Hammond WS. The effect of prescription size on acquisition of maintenance medications. J Gen Intern Med. 1993;8(6):306-310.

4. Carroll NV, Brusilovsky I, York B, Oscar R. Comparison of costs of community and mail service pharmacy. J Am Pharm Assoc (2003). 2005;45(3):336-343.

5. Millar J, MacKinnon W, Struthers MV, Vass C. A pilot study to investigate the use of installment dispensing as a method of reducing drug wastage owing to adverse drug reactions. Br J Gen Pract. 2003;53(492):550-552.

6. Hawksworth GM, Wright DJ, Chrystyn H. A detailed analysis of day-to-day unwanted medicinal products returned to community pharmacists for disposal. J Soc Adm Pharm. 1996;13:215-222.

7. Walton SM, Arondekar BV, Johnson NE, Schumock GT. A model for comparing unnecessary costs associated with various prescription fill-quantity policies: illustration using VA data. Managed Care Pharm. 2001;7(5):386-390.

8. Federal supply schedule prices. US Department of Veterans Affairs website. Published 2013. Accessed August 10, 2013.

9. About the VA San Diego Healthcare System. US Department of Veteran Affairs website. Published 2013. Accessed August 7, 2013.

10. Health benefits: health benefits copay rates. US Department of Veteran Affairs website. Published 2013. Accessed August 8, 2013.

11. Federal benefits for veterans, dependents and survivors. US Department of Veterans Affairs website. Published 2013. Accessed August 8, 2013.

12. Domino ME, Olinick J, Sleath B, Leinwand S, Byrns PJ, Carey T. Restricting patients’ medication supply to one month: saving or wasting money? Am J Health Syst Pharm. 2004;61(13);1375-1379.

13. Parikh SP, Dishman BR, Smith TL. Ninety-day versus thirty-day drug-dispensing systems. Am J Health Syst Pharm. 2001;58(13);1190-1191.

14. Chan LL, Crumpler WL, Jacobson AK. Implementation of pharmacist-managed anticoagulation in patients receiving newer anticoagulants. Am J Health Syst Pharm. 2013;70(15):1285-1286,1288.

15. Khandelwal N, Duncan I, Ahmed T, Rubinstein E, Pegus C. Impact of clinical oral chemotherapy program on wastage and hospitalizations. J Oncol Practice. 2011;7(3 suppl):e25s-e29s.

16. Insights: adherence: new program decreases cancer drug waste. Express Scripts website. Published August 9, 2012. Accessed August 12, 2013.

17. McHorney CA, Victor Spain C, Alexander CM, Simmons J. Validity of the adherence estimator in the prediction of 9-month persistence with medications prescribed for chronic diseases: a prospective analysis of data from pharmacy claims. Clin Ther. 2009;31(11);2584-2607.