Cost and Utilization Outcomes Associated With a Dosage Consolidation and Quantity Limits Program

, , , ,
The American Journal of Pharmacy Benefits, October 2010, Volume 2, Issue 5

Dose consolidation and quantity limit initiatives decreased utilization and cost outcomes for select mental health drugs in a fee-for-service Medicaid environment.

Dose consolidation and quantity limitations (QLs) are 2 strategies used by pharmacy administrators and decision makers to assist in controlling drug costs and increased utilization per patient. Dose consolidation, also referred to as dose optimization, is a drug cost management intervention. In dose consolidation, regimens consisting of lower-strength products taken as multiple units per day are converted to an equivalent daily dose of a higher-strength tablet or capsule.1

Once-daily maintenance medications with multiple dosage strengths and parallel pricing are typically appealing options for dose consolidation programs. There is limited published literature on the clinical effectiveness of these programs, but they are desirable from a cost-savings standpoint. In addition, simplifying a patient’s drug regimen may decrease the chance of missed doses and increase adherence, while providing the same total daily dose.2

It is important to note that not all patients are candidates for dose consolidation. A physician may wish to continue a patient on the current regimen based on many factors. These include, but are not limited to, patient age, adverse events, tolerability, tablet or capsule size, dose stabilization, and possible varying dosing regimens.3 With a new dose consolidation program, it is important to offer physicians the option of continuing their patients on the current dosing regimen if needed. This can be done via a prior authorization (PA) process.

Essentially, a dose consolidation policy encourages pharmacy providers to dispense 1 larger strength of a drug product rather than 2 smaller strengths. In addition, QLs restrict pharmacy providers from dispensing more than the allowed tablet or capsule quantity per day. If the prescriber has a clinical reason for prescribing outside these limitations, a PA request is submitted to the payer before dispensing the written dose and/or quantity. If a PA is not received, the pharmacy must dispense the product within the limits established by the payer.

Certain mental health drugs such as aripiprazole, olanzapine, venlafaxine XR, escitalopram, and sertraline are typically dosed once daily, and are available in multiple strengths with very small differences in average wholesale price (AWP) from strength to strength. As such, these particular drug products represent an excellent target for a dose consolidation and/or QL program.


Our program was implemented within a fee-for-service state Medicaid environment. A 3-month claims review for select once-daily mental health drugs (aripiprazole, olanzapine, venlafaxine XR, escitalopram, and sertraline) from November 1, 2005, to January 1, 2006, was done to determine potential dose consolidation opportunities. Analysis of these data indicated that lower-strength forms of these products were frequently being prescribed as multiple units per day, despite the availability of higher-strength forms of the same product. In addition to economic concerns, the relative prevalence of this dosing inefficiency prompted concerns about increased pill burden for patients, added complexity of drug regimens, and a potentially negative impact on patient adherence to therapy.

As a result of the initial review, the state’s drug utilization review board voted in support of a combined dosage consolidation protocol and QLs for dispensing the lower strengths of these 5 medications. It is important to note that the highest strength of each product was not subject to any dose consolidation or QL initiatives. In addition to the highest doses, aripiprazole 2 mg, aripiprazole 20 mg, and olanzapine 15 mg also were not subject to these initiatives because higher practical consolidation doses were not available for these strengths (eg, 2 × 20 mg could not be consolidated to 1 × 40 mg).

Table 1

illustrates the exact QLs placed on each product and strength.

The QLs went into effect October 1, 2006, and limited prescribing to 1 unit (tablet or capsule) per day with select dosage strengths of these products. Two months before the implementation of this program, a formal patient-specific mailing and PA form were sent to 250 physicians of potentially affected patients.

The objective of this analysis was to assess the cost and utilization outcomes associated with this dose consolidation program a year after its original implementation date.


Clinical Education Outreach Based on a pharmacy claims assessment between May and July 2006, 418 patients were identified as receiving lower-strength forms of aripiprazole, olanzapine, venlafaxine XR, escitalopram, and sertraline being prescribed as multiple units per day when higher-strength, single-unit alternatives were readily available. In August 2006, formal patient-specific mailings were sent to 250 physicians outlining the background and objectives of the correspondence, as well as notifying the physicians of forthcoming implementation of QLs for these 5 drugs, effective October 1, 2006.

The mailing included a pharmacist-reviewed, patient-specific prescription profile populated with the most recent claims data. The profile identified each patient under that provider’s care who appeared to be receiving a lower-strength form of 1 of the above-mentioned medications in quantities greater than 1 unit per day. The mailing provided patient-specific recommendations as to a more appropriate consolidated dosage regimen for the prescriber to consider. Providers were asked to review each patient’s record to determine whether the patient’s dosing was appropriate for consolidation. The profile also identified the last pharmacy where the patient’s prescription had been filled, as well as the phone and fax numbers for that pharmacy to facilitate any change in regimen. The mailing also contained a physician response form that allowed clinical staff to collect information from the provider as to his or her anticipated patient-specific course of action in a follow- up to this initiative. Finally, the mailing included a PA request form, in the event that there was clinical justification for the patient to remain on his or her current dosage form and/or regimen.

Cost and Utilization

To assess the effects of this initiative on cost and utilization, 6 months of pharmacy claims data from April 1, 2007, through September 30, 2007 (postimplementation) were compared with 6 months of pharmacy claims data from April 1, 2006, through September 30, 2006 (preimplementation). Data on prescription volume, ingredient cost per unit based on discounted AWP, total units, and total days of therapy were retrieved for each strength of the selected mental health products per month. Subsequently, mean units per day (± standard deviation [SD]), mean units per prescription (±SD), mean ingredient cost per unit (±SD), and 6-month savings figures were calculated. The latter was obtained using the following equation: (postimplementation units per day — preimplementation units per day) × postimplementation average cost per unit × postimplementation total days of therapy

In addition, we determined mean totals (±SD) for each product regardless of strength, as well as cumulative mean totals (±SD) (all drugs, all strengths).

Statistical Analysis

Multifactor analysis of variance with Student-Newman-Keuls post-hoc testing was used to determine whether significant differences existed between the preimplementation and postimplementation means. Results were considered to be statistically significant if the observed level of significance (P value) was less than .05. All analyses were performed using NCSS 2000.4


Education Outreach

Of 250 total mailings, physician response forms were received from 136 physicians, for a 54.4% active response rate. Of the response forms received, 70 (51%) responders indicated they would review therapy with the patient by phone, by appointment, or at the next scheduled office visit. Another 32 (24%) responders chose not to intervene in therapy, and 27 (20%) responders indicated that it was not their patient or that they were no longer treating the patient. The remaining 7 (5%) responders did not indicate a reason, choosing “other” on the response form. See

Table 2

for a full breakdown of physician mailing responses. Physicians were asked to rate the usefulness or value of the information provided in the mailing. Based on a scale of 1 to 10 (not useful at all to extremely useful), the average weighted score was 5.4 out of 10.

Utilization Assessment

The overall assessment for all drugs and strengths showed a statistically significant decrease in mean units per day, mean units per prescription, and mean ingredient cost per unit (—0.10 ± 0.28, P <.001; —2.19 ± 10.30, P <.001; and —$0.06 ± $5.94, P <.001, respectively) in the postimplementation group compared with the preimplementation group (

Table 3


The data were further evaluated for statistical significance with respect to mean units per day and mean units per prescription for each product and strength. Overall product evaluation (all strengths) showed a statistically significant decrease in both mean units per day and mean units per prescription with aripiprazole (-0.04 ± 0.07, P<.001, and -1.07 ± 5.92, P <.01, respectively), venlafaxine XR (-0.20 ± 0.14, P <.001, and -7.13 ± 4.66, P <.001, respectively), sertraline (-0.18 ± 0.14, P <.001, and -4.03 ± 3.71, P <.001, respectively), and olanzapine (-0.21 ± 0.23, P <.001, and -5.75 ± 8.55, P <.001, respectively). The overall escitalopram group had a decrease in mean units per day and mean units per prescription; however, only the latter was statistically significant (-1.86 ± 4.33, P <.01). Sertraline (as a branded product) had an increase in utilization for both data points (0.10 ± 0.53 and 6.17 ± 16.78, respectively), neither of which was statistically significant.

An evaluation also was performed on each of the individual drug strengths. For aripiprazole, a statistically significant decrease was noted in both utilization measures for the 5-mg, 10-mg, and 15-mg doses (P <.05). Aripiprazole 2 mg showed an increase in both utilization measures, but only the mean units per day were statistically significant (P <.05). Aripiprazole 20 mg and 30 mg had statistically significant increases in mean units per prescription (P <.05). Changes in mean units per day for these 2 doses were not statistically significant.

In the venlafaxine XR group, the 37.5-mg and 75-mg strengths had a statistically significant decrease in both utilization measures. The 150-mg strength showed a slight decrease (-0.01 ± 0.02) in mean units per day and a slight increase (0.01 ± 0.86) in mean units per prescription, but neither was statistically significant.

In the escitalopram group, the 5-mg and 10-mg doses showed a statistically significant decrease in mean units per prescription (P <.05). The 20-mg dose showed an increase in both utilization measures, neither of which was statistically significant.

During the course of our preimplementation and postimplementation analyses, branded sertraline lost its patent protection and a generic equivalent was launched. In the generic sertraline group, all doses had a decrease in both utilization measures. However, changes in mean units per day in the 100-mg strength were not statistically significant. In the branded sertraline group, none of the changes in utilization measures for each dose were statistically significant.

Lastly, in the olanzapine group, the 2.5-mg, 5-mg, 10- mg, and 15-mg strengths all showed a statistically significant decrease for both utilization measures (P <.05). Olanzapine 7.5 mg did show a decrease in both measures, although neither was statistically significant. The 20-mg strength showed a decrease in mean units per day and an increase in mean units per prescription; the latter was statistically significant (P <.05).

eAppendix Tables A-F

have a complete breakdown of results for each medication and are available at


Cost Outcomes and Savings

Overall, these efforts resulted in a statistically significant decrease in mean ingredient cost per unit (-$0.06 ± $5.94, P <.001) when comparing the postimplementation group with the preimplementation group. When comparing postimplementation with preimplementation values for each of the individually targeted drug products, all showed a statistically significant increase in mean ingredient cost per unit (P <.001) secondary to intervention, with the exception of sertraline (-$1.95 ± $0.02, P <.001).

Using the equation shown in the Methods section, an overall 6-month savings of $310,679.46 was calculated. This translated to per member per month cost savings of approximately $0.37. The overall Zoloft group (all strengths) was the only one not to show a cost savings (-$935.60). This, of course, may be attributable to the substantial decrease in brand Zoloft use secondary to the launch of generic sertraline, as well as manufacturer pricing changes that occurred in anticipation of this lost market share. Complete cost savings data are in

Table 4



It is difficult to determine whether the clinical education and notifications sent to physicians in August 2006 played a role in any of the cost and utilization changes shown in this analysis. These notifications were to alert physicians to possibly affected patients and offer them information on PA in the event that there was clinical justification for continuing a patient on his or her current dose. Although there were only 136 responses, 51% of the physicians who responded mentioned that they would write a new prescription for a once-daily dosage regimen at a higher strength.

Our analysis was designed to examine the effects of a consolidation and QL program on utilization and cost. Unfortunately, we did not simultaneously test for clinical or humanistic outcome measures in this population. Although such a study would add value, it was not the intention of our analysis. In addition, assessing clinical or humanistic measures in this population would be extremely difficult because of cost, time, manpower, and the type of data available to us (pharmacy claims data only). Additional research assessing QLs and their effects on clinical outcomes are warranted for validation of these implementation programs.

It is important to note that Zoloft’s generic equivalent sertraline became available in the third quarter of 2006, which occurred during our preimplementation time frame. It also is important to note the vast difference in AWP cost when sertraline became available in our preimplantation phase compared with its cost in the postimplementation phase. This large difference in unit cost is most likely attributable to the generic exclusivity of sertraline, which limits the production of other generic forms of sertraline for the first 180 days. During our postimplementation phase, other generic forms were available, therefore drastically lowering the cost.

It is also difficult to truly assess whether, and to what extent, these events affected our data and our analysis. Our units per day and units per prescription seemed not to be affected because they were calculated independent of cost and based on total days of therapy and total prescriptions. On the other hand, effects on cost were very apparent in the sertraline group. Ingredient costs per unit decreased by an average of $1.95, the largest change witnessed in our cost analysis.

Savings for the 6-month postimplementation evaluation were calculated based on units per day and discounted AWP cost per unit; additional rebate discounting was not taken into consideration. Although the study showed a cost savings of $310,679.46 ($621,358.92 annualized) during the 6-month postimplementation period, the study did not assess additional administrative costs or funds tied to program implementation or rebate implications (if any).

The Zoloft group was the only one to show an increase in cost. Although it is not certain, we believe the substantial drop in brand Zoloft usage secondary to generic introduction and manufacturer price increases on the branded product likely contributed to the confounded Zoloft figures. On the other hand, introduction of multiple generic forms of sertraline before our postimplementation phase resulted in cost savings of $9317.32, which helped offset the increase in the Zoloft group. Assessment of PA requests and subsequent results to examine this potential relationship may be warranted in subsequent studies.


Overall, dose consolidation and QL initiatives were successful in positively affecting daily dosing patterns and cost outcomes for these selected mental health products. Research focusing on clinical and humanistic outcomes in lieu of these strategies is needed to help further validate these programs.