Psychometric Approaches to Defining Cognitive Phenotypes in the Old Order Amish

OBJECTIVE

Memory and cognitive problems are central to the diagnosis of Alzheimer's disease (AD). Psychometric approaches to defining phenotypes can aid in identify genetic variants associated with AD. However, these approaches have mostly been limited to affected individuals. Defining phenotypes of both affected and unaffected individuals may help identify genetic variants associated with both AD and healthy aging. This study compares psychometric methods for developing cognitive phenotypes that are more granular than clinical classifications.

METHODS

682 older Old Order Amish individuals were included in the analysis. Adjusted Z-scores of cognitive tests were used to create four models including 1) global threshold scores or 2) memory threshold scores, and 3) global clusters and 4) memory clusters. An ordinal regression examined the coherence of the models with clinical classifications [cognitively impaired (CI), mildly impaired (MI), cognitively unimpaired (CU)], APOE-e4, sex, and age. An ANOVA examined the best model phenotypes for differences in clinical classification, APOE-e4, domain Z-scores (memory, language, executive function, and processing speed), sex, and age.

RESULTS

The memory cluster identified four phenotypes and had the best fit (χ² = 491.66). Individuals in the worse performing phenotypes were more likely to be classified as CI or MI and to have APOE-e4. Additionally, all four phenotypes performed significantly differently from one another on the domains of memory, language, and executive functioning.

CONCLUSIONS

Memory cluster stratification identified the cognitive phenotypes that best aligned with clinical classifications, APOE-e4, and cognitive performance We predict these phenotypes will prove useful in searching for protective genetic variants. This article is protected by copyright. All rights reserved.

Implications of controlling for comorbid conditions in cost-of-illness estimates: a case study of osteoarthritis from a managed care system perspective

OBJECTIVES

Current methods for estimating the cost of illness inconsistently control for the effect of comorbid conditions. This analysis examines the implications of controlling for comorbid conditions on the estimated cost of illness. These implications are illustrated using the cost of osteoarthritis as an example.

METHODS AND DATA

Medical claims data from 1996 were obtained for inpatient, outpatient, and pharmacy services for members in four United HealthCare health plans. Total annual costs for osteoarthritis (OA) were compared to costs among an equal number of comparison members. Multivariate regression analysis was used to compare the natural log of costs between the OA and comparison groups under two alternative controls for comorbid conditions: no controls, and controls for all conditions.

RESULTS

Controlling for no or all comorbid conditions resulted in estimates of the annual cost of members with OA that ranged between 261% and 151% of the cost of members without OA, respectively.

CONCLUSIONS

Existing cost-of-illness estimates may seriously underestimate the true cost by including statistical controls for all comorbid conditions, or seriously overestimate the true cost by failing to control for enough comorbid conditions. In the case of OA, the range of potential bias is substantial.

Postmarketing surveillance of medications and pregnancy outcomes: clarithromycin and birth malformations

PURPOSE

This retrospective surveillance study used linked administrative claims data and medical records to determine the rate and types of birth malformations in infants born to women exposed to the antibiotic, clarithromycin (Biaxin), during the first trimester of pregnancy.

METHODS

Pharmacy and hospital claims from eight geographically diverse health plans were used to identify women who had a delivery claim within 270 days of a clarithromycin prescription over a 5-year period (1991-1995). Hospital delivery admission medical records for 143 mothers and their 149 infants were abstracted to identify birth malformations.

RESULTS

Five infants were identified with major malformations, three with minor malformations, and four with undescended testicles likely to resolve with time. The observed rate of 3.4% (95% CI, 0.5, 6.3) for major malformations was not statistically significantly different compared to an expected rate of 2.8% based on earlier national data. There was no consistency across types of major malformations.

CONCLUSIONS

These results provide no evidence that clarithromycin is a likely major teratogen in humans. Use of claims data is one way to evaluate quickly and efficiently the safety of prescription medications in humans during pregnancy, especially when both exposure and outcome are rare.

One-year costs of second-line therapies for depression

BACKGROUND

We compared patterns of medical resource utilization and costs among patients receiving a serotonin-norepinephrine reuptake inhibitor (venlafaxine), one of the selective serotonin reuptake inhibitors (SSRIs), one of the tricyclic agents (TCAs), or 1 of 3 other second-line therapies for depression.

METHOD

Using claims data from a national managed care organization, we identified patients diagnosed with depression (ICD-9-CM criteria) who received second-line antidepressant therapy between 1993 and 1997. Second-line therapy was defined as a switch from the first class of antidepressant therapy observed in the data set within 1 year of a diagnosis of depression to a different class of antidepressant therapy. Patients with psychiatric comorbidities were excluded.

RESULTS

Of 981 patients included in the study, 21% (N = 208) received venlafaxine, 34% (N = 332) received an SSRI, 19% (N = 191) received a TCA, and 25% (N = 250) received other second-line antidepressant therapy. Mean age was 43 years, and 72% of patients were women. Age, prescriber of second-line therapy, and prior 6-month expenditures all differed significantly among the 4 therapy groups. Total, depression-coded, and non-depression-coded 1-year expenditures were, respectively, $6945, $2064, and $4881 for venlafaxine; $7237, $1682, and $5555 for SSRIs; $7925, $1335, and $6590 for TCAs; and $7371, $2222, and $5149 for other antidepressants. In bivariate analyses, compared with TCA-treated patients, venlafaxine- and SSRI-treated patients had significantly higher depression-coded but significantly lower non-depression-coded expenditures. Venlafaxine was associated with significantly higher depression-coded expenditures than SSRIs. However, after adjustment for potential confounding covariables in multivariate analyses, only the difference in depression-coded expenditures between SSRI and TCA therapy remained significant.

CONCLUSION

After adjustment for confounding patient characteristics, 1-year medical expenditures were generally similar among patients receiving venlafaxine, SSRIs, TCAs, and other second-line therapies for depression. Observed differences in patient characteristics and unadjusted expenditures raise questions as to how different types of patients are selected to receive alternative second-line therapies for depression.

Financial incentives and drug spending in managed care

Pharmaceutical costs have been rising dramatically since 1995, growing 16.6% in 1998 alone. This rate of increase is more than four times that of all health care spending. Employers, managed care organizations and consumers are looking anew for ways to stem these rising costs, without denying patients effective care. Therefore, this Issue Brief is especially timely because it investigates how patient copayments and financial incentives for physicians affect drug spending in managed care.

Medical resource use and cost of venlafaxine or tricyclic antidepressant therapy. Following selective serotonin reuptake inhibitor therapy for depression

OBJECTIVE

An analysis of administrative and claims data was performed to compare the resource use and costs to a managed-care organisation of venlafaxine, a serotonin and norepinephrine reuptake inhibitor (SNRI), versus tricyclic antidepressant (TCA) therapy, after switching from a selective serotonin reuptake inhibitor (SSRI).

DESIGN

One-year costs and frequencies of all medical services, and of services coded for depression, were compared between patients who received venlafaxine and TCA therapy as second-line therapy using bivariate and multivariate statistical analyses.

SETTING

Data were obtained from 9 individual health plans with more than 1.1 million covered lives affiliated with a national managed-care organisation.

PATIENTS AND PARTICIPANTS

Health plan members were included if they had a diagnosis of depression between July 1993 and February 1997. They also had to have at least 2 months of prescriptions for SSRI therapy followed by at least 2 months of venlafaxine or TCA therapy, and continuous enrollment in the plan from at least 6 months prior to 12 months following initiation of venlafaxine or TCA therapy. 188 patients who received venlafaxine and 172 patients who received TCAs met the inclusion criteria.

MAIN OUTCOME MEASURES AND RESULTS

Patients who received TCAs were slightly but significantly older (43 vs 40 years) than venlafaxine recipients and, during 6 months prior to initiating therapy, had significantly higher mean costs coded for depression ($US451 vs $US311) and costs not coded for depression ($US4500 vs $US2090). Psychiatrists prescribed a significantly higher proportion of venlafaxine than TCA prescriptions (46.3 vs 25.0%). Prior to adjusting for confounding characteristics, during 12 months following initiation of therapy, mean depression-coded costs were significantly higher for venlafaxine than TCA recipients ($US1948 vs $US1396) and mean costs not coded for depression were significantly lower ($US4595 vs $US6677). Overall costs were not significantly different ($US6543 for venlafaxine vs $US8073 for TCA). Significant cost differences were observed with primary care physicians as initial prescribers of second-line therapy but not with psychiatrists. However, costs between the 2 groups were similar after adjusting for confounding variables, including prior 6-month costs and initial prescriber of second-line therapy.

CONCLUSIONS

Payer costs are similar among patients receiving venlafaxine and TCA therapy following SSRI therapy. Higher costs of venlafaxine pharmacotherapy relative to TCA therapy may be offset by lower costs of other medical services. Differences in prescribing patterns and costs between primary care physicians and psychiatrists warrant further investigation.

Financial incentives and drug spending in managed care

This study estimates the impact of patient financial incentives on the use and cost of prescription drugs in the context of differing physician payment mechanisms. A large data set was developed that covers persons in managed care who pay varying levels of cost sharing and whose physicians are compensated under two different models: independent practice association (IPA)-model and network-model health maintenance organizations (HMOs). Our results indicate that higher patient copayments for prescription drugs are associated with lower drug spending in IPA models (in which physicians are not at risk for drug costs) but have little effect in network models (in which physicians bear financial risk for all prescribing behavior).

Hysterectomy vs. resectoscopic endometrial ablation for the control of abnormal uterine bleeding. A cost-comparative study

This study compared the costs of endometrial ablation using the uterine resectoscope to those of hysterectomy in a group of patients treated for abnormal uterine bleeding who were enrolled in a national managed health care organization. The cost of endometrial ablation during the periprocedural period was significantly lower than that of hysterectomy, with much of the difference coming from the hospitalization required for the latter procedure. The postprocedural cost for ablation was higher than for hysterectomy owing to the need for second ablations or hysterectomy in 13 of the 85 ablation patients. Preprocedure costs were not different between ablation and hysterectomy. A reanalysis of the data, however, that excluded patients who required a second ablation or hysterectomy suggested that these additional procedures were responsible for the higher postprocedural costs in the ablation group. Resectoscopic endometrial ablation for the treatment of abnormal uterine bleeding resulted in lower periprocedure costs and lower overall treatment costs to the health plan in the groups studied as compared with hysterectomy. Greater familiarity with the technique of resectoscopic endometrial ablation, improved patient selection for the procedure and the use of appropriate pharmacotherapy for suppressing endometrial growth prior to ablation probably substantially improve the rate of success, reduce postprocedural costs and further enhance the cost advantage of this procedure.

Patient adherence to prescribed potassium supplement therapy

We have investigated whether patient adherence ratios calculated from prescription refill data for potassium supplement medications differ depending on the type of supplement. By using automated pharmacy claims records from a large managed care organization, an index of adherence to prescribed therapy was calculated for each patient as a ratio of total days of drug supplied to the total number of days between prescription refills. The mean patient adherence to prescribed therapy ratios were compared among different potassium drug regimens. There were 2289 patients eligible for analysis; 65.9% were women, and the mean age was 57.6 years. The mean patient adherence ratio for one brand of extended-release tablet, K-DUR, was 0.81 (a majority of the patients were receiving 20 mEq/day). This was higher than the combined mean patient adherence ratio for all other supplements (0.73); the combined mean ratio for all other extended-release tablets (0.74); the combined mean ratio for all other tablets and capsules (0.74); the combined mean ratio for liquids (0.50); the combined mean ratio for liquids and powders (0.63); and equivalent to the ratio for another extended tablet, Micro-K (0.82). Regression analysis showed that increased patient adherence was seen among patients taking K-DUR tablets as compared with those taking other potassium supplements. Increased adherence among patients taking K-DUR remained statistically significant after controlling for number of prescriptions filled, dose, age, sex, and health plan location. Pharmacy claims data can be used effectively to measure patient adherence with potassium supplement therapy. Future research should relate patient adherence ratios to clinical outcomes.

Using claims data for epidemiologic research. The concordance of claims-based criteria with the medical record and patient survey for identifying a hypertensive population

In this study, a method was developed to identify health plan members with hypertension from insurance claims, using medical records and a patient survey for validation. A sample of 2,079 patients from two study sites with medical service or pharmacy claims indicating a diagnosis of essential hypertension were surveyed, and the medical records of 182 of the 1,275 survey respondents were reviewed. Where the criteria to identify hypertensive patients used both the medical and pharmacy claims, there was 96% agreement with either the medical record or the patient survey. Where the criteria relied on medical claims alone, the agreement rate decreased to 74% with the medical record and 64% with the patient survey. Where the criteria relied on the pharmacy claims alone, the agreement rate was 67% with the medical record and 75% with the patient survey. Combined evidence from medical service and pharmacy claims yielded a high level of agreement with alternative, more costly sources of data in identifying patients with essential hypertension. As it is more thoroughly investigated, claims data should become a more widely accepted resource for epidemiologic research.