Bryan Dowd

The impact of the 1987 federal regulations on the use of psychotropic drugs in Minnesota nursing homes.

OBJECTIVES The purpose of this study was to examine prevalence rates of psychotropic drug use by elderly nursing home residents 3 years before and 1 year after implementation of the 1987 Omnibus Budget Reconciliation Act drug regulations throughout the United States on October 1, 1990. METHODS A cohort study was conducted of elderly nursing home residents, for each of 4 study years (approximately 33,000 residents per year), of all nursing homes (n = 372) in Minnesota certified by Medicare and Medicaid. Data included (1) health status assessment and psychotropic drug use; (2) nursing home and care characteristics; and (3) county geographic and population characteristics. RESULTS Annual rates of antipsychotic drug use declined by one third over the 4-year period (23%, 22%, 19%, and 15% from 3 years before enforcement of the regulations to 1 year afterward). All differences were statistically significant. Antianxiety use rates were 11%, 12%, 12%, and 12%, respectively, and antidepressant use rates were 14%, 15%, 16%, 16%, respectively, for the 4 years. The latter two classes of drugs were not affected directly by the regulations. CONCLUSIONS Declines in the rates of antipsychotic drug use appear to be associated with anticipation of the regulations the year before and as the result of the regulations the year after the October 1990 implementation. A hypothesized medication shift to benzodiazepine drugs was not observed.

Impact of inappropriate drug use on health services utilization among representative older community-dwelling residents

BACKGROUND There is limited objective information regarding the impact of drugs identified as inappropriate by drug utilization review (DUR) or the Beers drugs-to-avoid criteria on health service use. OBJECTIVE The goal of this study was to examine the predictive validity of DUR and the Beers criteria employed to define inappropriate drug use in representative community residents, aged >or=68 years, as determined by the relationship of these criteria to health service use in older community residents. METHODS Data came from participants in the Duke University Established Populations for Epidemiologic Studies of the Elderly seen in 1989/1990 and for whom information was also available 3 years later. Two sets of inappropriate drug use criteria were examined: (1) DUR regarding dosage, duration, duplication, and drug-drug and drug-disease interactions; and (2) the Beers criteria, applied to drug use reported in an in-home interview. Outpatient visits and nursing-home entry were determined by personal report; hospitalization information came from Medicare Part A files from the Centers for Medicare and Medicaid Services. RESULTS A total of 3165 participants were available at the fourth interview in 1989/1990. The majority were aged >74 years (51.1%), white (64.8%), women (64.7%), had fair or poor health (77.0%), consistently saw the same physician (86.9%), and possessed supplemental health insurance (62.8%). Use of inappropriate drugs meeting DUR criteria, especially for drug-drug or drug-disease interaction problems, was associated with increased outpatient visits (P<0.05) but not with time to hospitalization or time to nursing home entry. The use of inappropriate drugs according to the Beers criteria was associated with reduced time to hospitalization (adjusted hazard ratio, 1.20; 95% CI, 1.04-1.39) but not to outpatient visits or nursing home entry. CONCLUSIONS Our data suggest that in representative community residents aged >or=68 years, current criteria for inappropriate drug use should be used with caution in evaluating quality of care because they have minimal impact on use of health services. We found increases only in the use of outpatient services (with DUR) and more rapid use of hospitalization (with the Beers criteria).

The effect of premiums on the small firm's decision to offer health insurance

Many small firms (fewer than 50 employees) do not offer health insurance. We investigated the role of premiums in the small firms decision to offer health insurance, using data from a 1993 survey of 2,000 small firms in Minnesota. Selectively-corrected equations were estimated to predict the premiums faced by firms offering and not offering insurance. The elasticity of demand for health insurance, calculated at the mean of the data, was -3.91 for single coverage and -5.82 for family coverage. We contrast these results to the much lower responsiveness found in experimental studies and suggest that our findings are more likely to model the small firms demand for health insurance.

How does the culture of medical group practices influence the types of programs used to assure quality of care

Objective: It is widely acknowledged that the culture of medical group practices greatly influences the quality of care, but little is known about how cultures are translated into specific types of programs focused on quality. This study explores this issue by assessing the influence of the organizational culture on these types of programs in medical group practices in the upper Midwest. Design and Methods: Data were obtained from two surveys of medical group practices. The first survey was designed to assess the culture of the practice using a nine-dimension instrument developed previously. The second survey was designed to obtain organizational structure data including the programs identified by the literature as important to the quality of care in medical practices. Completed surveys were obtained from eighty-eight medical groups. The relationship of the group practice culture to structural programs focused on quality of care was analyzed using logistic regression equations. Results: Several interesting patterns emerged. As expected, practices with a strong information culture favor electronic data systems and formal programs that provide comparative or evidence-based data to enhance their clinical practices. However, those with a quality-centered culture appear to prefer patient satisfaction surveys to assess the quality of their care, while practices that are more business-oriented rely on bureaucratic strategies such as benchmarking and physician profiling. Cultures that emphasize the autonomy of physician practice were negatively (but not at a statistically significant level) associated with all the programs studied. Practices with a highly collegial culture appear to rely on informal peer review mechanisms to assure quality rather than any of the structural programs included in this analysis. Conclusion: This study suggests that the types of quality programs that group practices develop differ according to their cultures. Consequently, it is important for practice administrators and medical directors to develop quality assurance programs that fit their cultures if they are to gain buy-in by their clinicians. Future research should assess the effect of culture-structure fit on quality and safety outcomes.

The Determinants of Time off Work after Childbirth

Relatively little is known about the role that leave policies--family, parental, or maternity-leave policies--play in facilitating time off work after childbirth. Yet time off is a critical element of leave policies, as it facilitates the mothers recovery from childbirth and promotes maternal-infant attachment. Using data from Minnesota, the state with the highest rate of female labor force participation, we examine the extent to which policies, relative to personal, job, and workplace characteristics, determine the duration of womens childbirth-related leaves from work. A random sample of women identified from vital statistics records is used to estimate the relationship between leave policies and time off work after childbirth. Of our sample 85 percent had access to some paid leave benefits, although only 46 percent had paid maternity leave benefits. The difference in duration of leave between women with and without paid leave policies was approximately four weeks, a substantial difference for most women and their infants. Paid leave policies and spousal earnings as primary determinants of maternal time off work, suggest problems in the use of unpaid leave for economically vulnerable women.

Maternity leave duration and postpartum mental and physical health: Implications for leave policies

This study examines the association of leave duration with depressive symptoms, mental health, physical health, and maternal symptoms in the first postpartum year, using a prospective cohort design. Eligible employed women, eighteen years or older, were interviewed in person at three Minnesota hospitals while hospitalized for childbirth in 2001. Telephone interviews were conducted at six weeks (N = 716), twelve weeks (N = 661), six months (N = 625), and twelve months (N = 575) after delivery. Depressive symptoms (Edinburgh Postnatal Depression Scale), mental and physical health (SF-12 Health Survey), and maternal childbirth-related symptoms were measured at each time period. Two-stage least squares analysis showed that the relationship between leave duration and postpartum depressive symptoms is U-shaped, with a minimum at six months. In the first postpartum year, an increase in leave duration is associated with a decrease in depressive symptoms until six months postpartum. Moreover, ordinary least squares analysis showed a marginally significant linear positive association between leave duration and physical health. Taking leave from work provides time for mothers to rest and recover from pregnancy and childbirth. Findings indicate that the current leave duration provided by the Family and Medical Leave Act, twelve weeks, may not be sufficient for mothers at risk for or experiencing postpartum depression.

Computation of standard errors.

OBJECTIVES We discuss the problem of computing the standard errors of functions involving estimated parameters and provide the relevant computer code for three different computational approaches using two popular computer packages. STUDY DESIGN We show how to compute the standard errors of several functions of interest: the predicted value of the dependent variable for a particular subject, and the effect of a change in an explanatory variable on the predicted value of the dependent variable for an individual subject and average effect for a sample of subjects. EMPIRICAL APPLICATION: Using a publicly available dataset, we explain three different methods of computing standard errors: the delta method, Krinsky–Robb, and bootstrapping. We provide computer code for Stata 12 and LIMDEP 10/NLOGIT 5. CONCLUSIONS In most applications, choice of the computational method for standard errors of functions of estimated parameters is a matter of convenience. However, when computing standard errors of the sample average of functions that involve both estimated parameters and nonstochastic explanatory variables, it is important to consider the sources of variation in the functions values.

Risk Factors Associated With Cognitive, Functional, and Behavioral Trajectories of Newly Diagnosed Dementia Patients

Background: Dementia results in changes in cognition, function, and behavior. We examine the effect of sociodemographic and clinical risk factors on cognitive, functional, and behavioral declines in incident dementia patients. Methods: We used longitudinal data from the National Alzheimer’s Coordinating Center to evaluate cognitive (Mini-Mental State Exam [MMSE]), functional (Functional Activities Questionnaire [FAQ]), and behavioral (Neuropsychiatric Inventory Questionnaire [NPI-Q] severity score) trajectories for incident dementia patients over an 8-year period. We evaluated trajectories of 457 patients with mixed effects linear regression models. Results: In the first year, cognition worsened by −1.518 (95% confidence interval [CI] −1.745, −1.291) MMSE points (0–30 scale). Education, race, and region of residence predicted cognition at diagnosis. Age of onset, geographic region of residence, and history of hypertension and congestive heart failure predicted cognitive changes. Function worsened by 3.464 (95% CI 3.131, 3.798) FAQ points in the first year (0–30 scale). Cognition, gender, race, region of residence and place of residence, and a history of stroke and hypercholesterolemia predicted function at diagnosis. Place of residence and a history of diabetes predicted functional changes. Behavioral symptoms worsened by 0.354 (95% CI 0.123, 0.585) NPI-Q points in the first year (0–36 scale). Age of onset, region of residence, and history of hypertension and psychiatric problems predicted behaviors at diagnosis. Cognition explained changes in behavior. Conclusions: Sociodemographic characteristics and clinical comorbidities predict cognitive and functional changes. Only cognitive status explains behavioral decline. Results provide an understanding of the characteristics that impact cognitive, functional, and behavioral decline.

Health plan choice in the twin cities medicare market

This paper examines the relationship between characteristics of Medicare beneficiaries and their choice of health plan in the Twin Cities during 1988. This analysis provides the first comparison of beneficiaries in the basic fee-for-service (FFS) Medicare sector (without a supplementary policy) to beneficiaries in the FFS sector with a supplementary policy, enrollees in independent practice associations (IPAs), and network health maintenance organizations (HMOs). The site and time period are important because there were five large, mature HMOs with TEFRA-risk contracts operating at that time, enrolling 50% of Medicare beneficiaries in the market area. We find that the oldest, poorest and, to a lesser extent, the sickest Medicare beneficiaries were most likely to have basic FFS Medicare coverage without supplementary insurance. The youngest enrollees are found in network HMOs. The availability of group coverage and premium subsidies are positively associated with choice of FFS with a supplementary policy. Government policy concerning Medicare HMO premiums appears to contribute to the poorest beneficiaries facing the highest out-of-pocket costs.

How effective is health coaching in reducing health services expenditures

Background:Health coaching interventions aim to identify high-risk enrollees and encourage them to play a more proactive role in improving their health, improve their ability to navigate the health care system, and reduce costs. Objectives:Evaluate the effect of health coaching on inpatient, emergency room, outpatient, and prescription drug expenditures. Research Design:Quasiexperimental pre-post design. Health coaching participants were identified over the 2-year time period 2009–2010. Propensity scores facilitated matching eligible participants and nonparticipating controls on a one-to-one basis using nearest kernel techniques. Difference in differences logistic and generalized linear models addressed the impact of health coaching on the probability of incurring costs and levels of inpatient, emergency room, outpatient, and prescription drug expenditures, respectively. Measures:Administrative claims data were used to analyze health services expenditures preparticipation and post health coaching participation time periods. Results:Of the 6940 health coaching participants, 1161 participated for at least 4 weeks and had a minimum of 6 months of claims data preparticipation and postparticipation. Although the probability of incurring costs and expenditure levels for emergency room services were not affected, the probability of incurring inpatient expenditures and levels of outpatient and total costs for health coaching participants fell significantly from preparticipation to postparticipation relative to controls. Estimated outpatient and total cost savings were