The social drug lag: an examination of pharmaceutical approval delays in Medicaid formularies

Making a case for employing a societal perspective in the evaluation of Medicaid prescription drug interventions

The choice of a perspective is among the most critical influences of the potential outcome of an economic evaluation, since it determines whose interest is relevant in any given analysis. For publicly funded programmes such as Medicaid, and now Medicare, it is important that economic evaluations are undertaken from a societal perspective because such evaluations inform decisions about allocation of healthcare resources. It has been argued that approaches other than the societal perspective selectively include specific costs, while ignoring other costs that are very much more 'real', and hence lack theoretical foundation in welfare economics. In view of the importance of perspectives in economic evaluations, this paper reviews 25 existing reports of economic evaluations of interventions involving prescription drugs in the Medicaid programme to examine the perspectives employed in such evaluations, based on the specific cost and benefit measurements. No explicit statement of the perspective employed was included in any of the articles selected for this review. Based on an analysis of the cost measures, none of the studies were found to have adopted a societal perspective in their evaluation. Most studies were from the perspective of Medicaid as the payer and as such did not include costs and benefits from outside the Medicaid system. Ten of the identified evaluations of interventions focused just on costs related to prescription drugs. Six studies included an evaluation of the impact of the intervention on overall programme costs along with the costs of prescription drugs. The nine remaining evaluations employed a broader approach to include related effects of the drug-benefit intervention on costs and utilization of other healthcare services such as physician, outpatient and inpatient services. This review emphasizes the importance of a societal approach in evaluating the effects of interventions in Medicaid and other publicly funded drug benefit programmes. Existing evaluations fall short of employing such a broad perspective. This, along with the limitations in design and data, make findings from these studies less reliable than should be used to make major decisions regarding allocation of tax dollars. While methodological challenges to such an approach are valid and understandable, there is an increasing need to attempt evaluations of cost-containment strategies from a broad-based societal perspective to ensure continuity and sustainability of publicly funded drug benefit programmes such as Medicaid and Medicare.

Systemwide effects of Medicaid retrospective drug utilization review programs

Aggregate pooled cross-sectional and time-series annual state data for 1985 to 1992 were used to estimate the systemwide effects of retrospective drug utilization review programs (Retro-DUR) on Medicaid drug and nondrug outcomes. The results provide evidence that these programs produce significant cost savings in the drug budget without spillover effects (positive or negative) in other nondrug budgets within the Medicaid system. We also examine the influence of restricted formularies in this post-Retro-DUR era on drug and nondrug budgets in the Medicaid system; we find significant cost savings in the former but positive spillover effects in the latter.

What factors influence the prescribing of antidepressant pharmacotherapy? An assessment of national office-based encounters

OBJECTIVE

This study was designed to identify: 1) predictors of antidepressant pharmacotherapy among patients diagnosed with depression; and 2) predictors of prescription for either a selective-serotonin reuptake inhibitor (SSRI), or a serotonin-norepinephrine reuptake inhibitor (SNRI).

METHOD

Data from the 1995 National Ambulatory Medical Care Survey (NAMCS) were used to discern the number of office-based encounters documenting a diagnosis of depression (ICD-9-CM codes 296.2-296.36; 300.4; or 311) among patients eighteen years of age or older. Logistic regression-derived odds ratios (OR) and 95 percent confidence intervals (CI) were used to elucidate factors predictive of receipt of antidepressant pharmacotherapy, and, more specifically, factors predictive of receipt of an SSRI or an SNRI. Model variables included age (18-49 years as compared to > or = 50 years); race (white as compared to nonwhite, inclusive of Hispanics); gender; self-report of depression as a reason for the office-based encounter; and payer type (private insurance program as compared to public).

RESULTS

Among the estimated 18,046,293 office-based visits resulting in a diagnosis of depression, 56.2 percent of patients self-reported depression as a reason for the office-based encounter; 67.5 percent were prescribed or continued a regimen of antidepressant pharmacotherapy; and 48.3 percent were prescribed an SSRI or an SNRI. Factors predictive of receipt of antidepressant pharmacotherapy included age less than fifty years (OR = 1.30, CI = 1.01-1.67); female gender (OR = 1.45, CI = 1.13-1.85); and self-report of depression as a reason for the office-based encounter (OR = 1.98, CI = 1.57-2.51). Factors predictive of receipt of an SSRI or an SNRI included age less than fifty years (OR = 1.31, CI = 1.03-1.65); female gender (OR = 1.55, CI = 1.23-1.95); and self-report of depression as a reason for the office-based encounter (OR = 1.56, CI = 1.25-1.95). In addition, having private insurance increased the likelihood of having been prescribed an SSRI or SNRI by 46 percent (OR = 1.46, CI = 1.13-1.89).

CONCLUSIONS

Among patients with a diagnosis of depression, the pattern of prescribing antidepressant pharmacotherapy is influenced by a patient's age, gender, self-report of depression, and type of insurance coverage. Further research is required to discern the reasons for these observed effects and to advance clinically rational and equitable access to pharmacotherapeutic innovation.

Is there a U.S. drug lag? The timing of new pharmaceutical approvals in the G-7 countries and Switzerland

To test whether the United States is slow in approving new drugs, the date of approval of 34 important pharmaceuticals approved in the United States between 1970 and 1988 were compared with those of the other G-7 countries plus Switzerland. The drugs studied in the analysis were designated as having been especially important and therapeutically significant at the time of their approval by panels of physicians and pharmacists in the United States and France. Drug lags were measured for each country for each drug by computing the number of months elapsed from the time a drug was first approved by one of the countries in this sample. Although Switzerland was found to approve drugs generally more quickly than other countries, differences between countries tended to be small. The United States was relatively fast in approving the drugs, countering the contention that it suffers from a substantial drug lag.

Pharmacy-enforced outpatient drug treatment protocols: a case study of Medi-Cal restrictions for cefaclor

OBJECTIVE

To evaluate whether a pharmacy-enforced treatment protocol successfully limited the use of a high-cost medication to high-risk patients.

DESIGN

A case study cost-effectiveness analysis was conducted to evaluate a treatment protocol for cefaclor. Episodes of care were defined, healthcare expenditures for all services were aggregated, and demographic data were retrieved from a five percent random sample of California Medicaid (Medi-Cal) recipients. Data were available for episodes occurring before cefaclor was made available under Medi-Cal.

SETTING

Medi-Cal added cefaclor to its formulary, limiting its use to patients over 50 years of age with lower respiratory tract infections (LRTIs). The unit of analysis was an episode of outpatient antibiotic treatment.

PATIENTS

Confirmed LRTI episodes and unconfirmed LRTI cefaclor episodes were analyzed, including multiple episodes of treatment for individual patients. A total of 7855 non-cefaclor LRTI episodes and 2556 cefaclor episodes were analyzed.

MAIN OUTCOME MEASURES

The primary outcome measures were healthcare expenditures three months after the initiation of antibiotic therapy, differentiated by type of service.

RESULTS

Physicians directed cefaclor toward higher-risk patients over age 50 years, even in unconfirmed LRTI episodes. Cefaclor use was estimated to reduce posttreatment costs by $388 per patient (p < 0.001), primarily because of reduced hospital expenditures of $366 (p < 0.001).

CONCLUSIONS

Pharmacy-enforced outpatient drug treatment protocols may be a viable alternative to restrictive formularies and prior authorization. In the case of cefaclor, the Medi-Cal treatment protocol appeared to allow high-risk patients better access to a high-cost medication while reducing total posttreatment costs.

Utility of a transdermal delivery system for antihypertensive therapy. Part 1

A retrospective evaluation of patient-level Medicaid claims data from two states was undertaken to discern the fiscal utility of transdermally delivered clonidine versus both the oral formulation of clonidine and oral formulations of eight other antihypertensive agents. In the first phase of our two-part study, we compared paid claims data (n = 1,135) from Florida for transdermal and oral clonidine. Multivariate regression analysis was used to evaluate the incremental impact of six variables on health-care expenditures in the first year after patients were given a diagnosis of hypertension. These variables were: age, gender, prior utilization of medical services, regimen complexity, and dosage formulation. Patients prescribed transdermal clonidine experienced a significant (p less than or equal to 0.001) increase in prescription expenditures and significant reductions in the use of physician (p less than or equal to 0.05), laboratory (p less than or equal to 0.10), and hospital (p less than or equal to 0.05) services. Moreover, savings were maximized (p less than or equal to 0.001) where multi-drug regimens incorporated the transdermal delivery system. In the second phase of our study we compared paid claims data (n = 8,894) from South Carolina for transdermal clonidine and for nine oral antihypertensive agents: atenolol, captopril, clonidine, diltiazem, enalapril, metoprolol, prazosin, terazosin, and verapamil-SR. Once again, regression analysis was used, this time to evaluate the incremental impact of five variables on health-care expenditures in the first year post diagnosis: age, gender, prior utilization of medical services, regimen complexity, and Medication Possession Ratio (MPR), an index of compliance. The data from part 2 of our study revealed that patients assigned a b.i.d. oral antihypertensive agent experienced a significant reduction (p less than or equal to 0.05) in MPR and a significant (p less than 0.05) increase in health-care expenditures when compared to patients prescribed the transdermal delivery system and to patients prescribed once-daily oral medications. These data confirm previous findings concerning the impact of complicated dosing regimens on compliance in hypertensive patients. In this two-part paper we report the data from both phases of our study.