Recurrence after hospitalization for acute coronary syndrome among HIV-infected and HIV-uninfected individuals.

OBJECTIVES: We evaluated the association of HIV infection and immunodeficiency with acute coronary syndrome (ACS) recurrence, and with all-cause mortality as a secondary outcome, after hospitalization for ACS among HIV-infected and HIV-uninfected individuals. METHODS: We conducted a retrospective cohort study within Kaiser Permanente Northern California of HIV-infected and HIV-uninfected adults discharged after ACS hospitalization [types: ST-elevation myocardial infarction (STEMI), non-STEMI, or unstable angina] during 1996-2010. We compared the outcomes of ACS recurrence and all-cause mortality within 3 years, both overall by HIV status and stratified by recent CD4 count, with HIV-uninfected individuals as the reference group. Hazard ratios (HRs) were obtained from Cox regression models with adjustment for age, sex, race/ethnicity, year, ACS type, smoking, and cardiovascular risk factors. RESULTS: Among 226 HIV-infected and 86 321 HIV-uninfected individuals with ACS, HIV-infected individuals had a similar risk of ACS recurrence compared with HIV-uninfected individuals [HR 1.08; 95% confidence interval (CI) 0.76-1.54]. HIV infection was independently associated with all-cause mortality after ACS hospitalization overall (HR 2.52; 95% CI 1.81-3.52). In CD4-stratified models, post-ACS mortality was higher for HIV-infected individuals with CD4 counts of 201-499 cells/µL (HR 2.64; 95% CI 1.66-4.20) and < 200 cells/µL (HR 5.41; 95% CI 3.14-9.34), but not those with CD4 counts ≥ 500 cells/µL (HR 0.67; 95% CI 0.22-2.08), compared with HIV-uninfected individuals (P trend < 0.001). CONCLUSIONS: HIV infection and immunodeficiency were not associated with recurrence of ACS after hospitalization. All-cause mortality was higher among HIV-infected compared with HIV-uninfected individuals, but there was no excess mortality risk among HIV-infected individuals with high CD4 counts.

Post-partum weight loss and glucose metabolism in women with gestational diabetes: the DEBI Study.

AIMS: Women with gestational diabetes are at high risk for developing diabetes; post-partum weight loss may reduce the risk of diabetes. We evaluated the association of post-partum weight change with changes in glucose, insulin and homeostasis model assessment of insulin resistance in a subsample (n = 72) of participants from Diet Exercise and Breastfeeding Intervention (DEBI), a randomized pilot trial of lifestyle intervention for women with gestational diabetes. METHODS: Glucose and insulin were measured fasting and 2 h after an oral glucose tolerance test at 6 weeks and 12 months post-partum. Women were categorized by weight change (lost > 2 kg vs. maintained/gained) between 6 weeks and 12 months post-partum. RESULTS: Compared with women who maintained or gained weight, women who lost > 2 kg experienced significantly lower increases in fasting glucose [age-adjusted means: 0.1 mmol/l (95% CI -0.03 to 0.3) vs. 0.4 mmol/l (95% CI 0.3-0.6); P < 0.01] and 2-h insulin [10.0 pmol/l (95% CI -56.9 to 76.9) vs. 181.2 pmol/l (95% CI 108.3-506.9); P < 0.01] and a significant reduction in 2-h glucose [-0.9 mmol/l (95% CI -1.4 to -0.3) vs. 0.3 mmol/l (95% CI -0.3 to 0.9); P < 0.01]. In multiple linear regression models adjusted for age, Hispanic ethnicity, medication use, meeting the Institute of Medicine's recommendations for gestational weight gain, breastfeeding and randomized group, a 1-kg increase in weight was significantly associated with increases in fasting and 2-h glucose (P < 0.05), but was not associated with insulin or homeostasis model assessment of insulin resistance. CONCLUSIONS: In women with gestational diabetes, modest post-partum weight loss may be associated with improvements in glucose metabolism.

Trends in annual incidence of methicillin-resistant Staphylococcus aureus (MRSA) infection in HIV-infected and HIV-uninfected patients.

We describe trends in incidence rates of methicillin-resistant Staphylococcus aureus (MRSA) in HIV-infected and HIV-uninfected patients enrolled in a large northern California Health Plan, and the ratio of MRSA to methicillin-susceptible S. aureus (MSSA) case counts. Between 1995 and 2010, 1549 MRSA infections were diagnosed in 14060 HIV-infected patients (11·0%) compared to 89546 MRSA infections in 6597396 HIV-uninfected patients (1·4%) (P = 0·00). A steady rise in MRSA infection rates began in 1995 in HIV-uninfected patients, peaking at 396·5 infections/100000 person-years in 2007. A more rapid rise in MRSA infection rates occurred in the HIV-infected group after 2000, peaking at 3592·8 infections/100000 in 2005. A declining trend in MRSA rates may have begun in 2008-2009. Comparing the ratio of MRSA to MSSA case counts, we observed that HIV-infected patients shouldered a greater burden of MRSA infection during most years of study follow-up compared to HIV-uninfected patients.

Breastfeeding and growth during infancy among offspring of mothers with gestational diabetes mellitus: a prospective cohort study.

BACKGROUND: Breastfeeding (BF) may protect against obesity and type 2 diabetes mellitus in children exposed to maternal diabetes in utero, but its effects on infant growth among this high-risk group have rarely been evaluated. OBJECTIVES: The objective of this study was to evaluate BF intensity and duration in relation to infant growth from birth through 12 months among offspring of mothers with gestational diabetes mellitus (GDM). METHODS: Prospective cohort of 464 GDM mother-infant dyads (28% White, 36% Hispanic, 26% Asian, 8% Black, 2% other). Weight and length measured at birth, 6-9 weeks, 6 months and 12 months. Categorized as intensive BF or formula feeding (FF) groups at 6-9 weeks (study baseline), and intensity from birth through 12 months as Group 1: consistent exclusive/mostly FF, Group 2: transition from BF to FF within 3-9 months and Group 3: consistent exclusive/mostly BF. Multivariable mixed linear regression models estimated adjusted mean (95% confidence interval) change in z-scores; weight-for-length (WLZ), weight-for-age and length-for-age. RESULTS: Compared with intensive BF at 6-9 weeks, FF showed greater increases in WLZ-scores from 6 to 9 weeks to 6 months [+0.38 (0.13 to 0.62) vs. +0.02 (-0.15 to 0.19); p = 0.02] and birth to 12 months [+1.11 (0.87 to 1.34) vs. +0.53 (0.37 to 0.69); p < 0.001]. For 12-month intensity and duration, Groups 2 and 3 had smaller WLZ-score increases than Group 1 from 6 to 9 weeks to 6 months [-0.05 (-0.27 to 0.18) and +0.07 (-0.19 to 0.23) vs. +0.40 (0.15 to 0.64); p = 0.01 and 0.07], and birth to 12 months [+0.60 (0.39 to 0.82) and +0.59 (0.33 to 0.85) vs. +0.97 (0.75 to 1.19); p < 0.05]. CONCLUSIONS: Among offspring of mothers with GDM, high intensity BF from birth through 1 year is associated with slower infant ponderal growth and lower weight gain.

Moderate intensity sports and exercise is associated with glycaemic control in women with gestational diabetes.

AIM: To assess the association of regular, unsupervised sports and exercise during pregnancy, by intensity level, with glycaemic control in women with gestational diabetes (GDM). METHODS: Prospective cohort study of 971 women who, shortly after being diagnosed with GDM, completed a Pregnancy Physical Activity Questionnaire assessing moderate and vigorous intensity sports and exercise in the past 3 months. Self-monitored capillary glucose values were obtained for the 6-week period following the questionnaire, with optimal glycaemic control defined≥80% values meeting the targets<5.3mmol/L for fasting and <7.8mmol/L 1-hour after meals. Logistic regression estimated the odds of achieving optimal control; linear regression estimated activity level-specific least square mean glucose, as well as between-level mean glucose differences. RESULTS: For volume of moderate intensity sports and exercise ([MET×hours]/week), the highest quartile, compared to the lowest, had significantly increased odds of optimal control (OR=1.82 [95% CI: 1.06-3.14] P=0.03). There were significant trends for decreasing mean 1-hour post breakfast, lunch and dinner glycaemia with increasing quartile of moderate activity (all P<0.05). Any participation in vigorous intensity sports and exercise was associated with decreased mean 1-hour post breakfast and lunch glycaemia (both P<0.05). No associations were observed for fasting. CONCLUSION: Higher volumes of moderate intensity sports and exercise, reported shortly after GDM diagnosis, were significantly associated with increased odds of achieving glycaemic control. Clinicians should be aware that unsupervised moderate intensity sports and exercise performed in mid-pregnancy aids in subsequent glycaemic control among women with GDM.

Exposure to barbiturates in utero and during childhood risk of intracranial and spinal cord tumors.

Barbiturate exposure during childhood was assessed from medical records of 237 children with intracranial and spinal cord tumors and 474 matched controls in a prepaid health plan. In utero exposure was also examined in a subset of 86 "cases" and 172 controls whose mothers were health plan members during pregnancy. No association of in utero exposure to barbiturates was found [odds ratio (O.R.) = 0.96, 95% confidence interval (C.I.) = 0.47, 1.94]. An association was noted for barbiturate use during childhood (O.R. = 1.80, 95% C.I. = 1.18, 2.74) but was reduced (O.R. = 1.41, 95% C.I. = 0.89, 2.21) when history of epilepsy was taken into account and was no longer significant. An apparent dose-response effect disappeared after adjustment for a history of epilepsy. Although barbiturate use for epilepsy due to preexisting brain tumors clearly explains some of the observed association, the small, residual risk prevents us from ruling out a possible carcinogenic effect of barbiturates. Further study of cohorts of adult as well as childhood users of barbiturates and other anticonvulsants is recommended.

Obesity, health services use, and health care costs among members of a health maintenance organization.

BACKGROUND: Obesity is an independent risk factor for a variety of chronic diseases and is therefore a potential source of avoidable excess health care expenditures. Previous studies of obesity and health care costs have used group level data, applying estimates of population-attributable risks to estimates of US total costs of care for each obesity-related disease. OBJECTIVE: To quantify the association between body mass index (BMI) and health services use and costs stratified by age and use source at the patient level, a level of detail not previously reported. METHODS: In 17,118 respondents to a 1993 health survey of members of a large health maintenance organization, we ascertained through computerized databases all hospitalizations, laboratory services, outpatient visits, outpatient pharmacy and radiology services, and the direct costs of providing these services during 1993. RESULTS: There was an association between BMI and annual rates of inpatient days, number and costs of outpatient visits, costs of outpatient pharmacy and laboratory services, and total costs (P < or = .003). Relative to BMI of 20 to 24.9, mean annual total costs were 25% greater among those with BMI of 30 to 34.9 (rate ratio, 1.25; 95% confidence interval, 1.10-1.41), and 44% greater among those with BMI of 35 or greater (rate ratio, 1.44; 95% confidence interval, 1.22-1.71). The association between BMI and coronary heart disease, hypertension, and diabetes largely explained these elevated costs. CONCLUSION: Given the high prevalence of obesity and the associated elevated rates of health services use and costs, there is a significant potential for a reduction in health care expenditures through obesity prevention efforts.

The cost of acquired immunodeficiency syndrome in northern California. The experience of a large prepaid health plan.

We studied the cost of acquired immunodeficiency syndrome (AIDS) in the Kaiser Permanente Medical Care Program (KPMCP), northern California region. We report the costs of care to the KPMCP and introduce an innovative application of survival methods to cost analysis. From the beginning of the AIDS epidemic in 1981 to the end of June 1987, 866 cases of AIDS were recorded among members of the KPMCP. Estimates of the costs of care of these patients were derived from comprehensive chart reviews of a random sample of 71 patients whose conditions were diagnosed from January 1984 through June 1987. Total mean lifetime costs per patient were $32,816 (median, $28,677), whereas the mean hospital per diem cost was $20,446 per patient. As more care was shifted to outpatient services overtime, overall costs dropped, despite marked increases in the cost of outpatient medications such as zidovudine. The overall estimate of cost compared closely with other estimates of the cost of care in San Francisco, Calif, and it is lower than estimates from elsewhere in the United States, probably because of the low proportion of cases associated with intravenous drug use and the well-developed social support networks available to patients with AIDS in the San Francisco Bay Area.

Usual dietary fat intake and insulin concentrations in healthy women twins.

OBJECTIVE: To evaluate the associations between the usual intake of dietary fats and insulin concentrations. Insulin concentrations and insulin resistance have been positively related to risk for NIDDM, obesity, hypertension, dyslipidemia, and coronary heart disease, yet little is known of the environmental risk factors for relative hyperinsulinemia. Insulin resistance can be induced by high-fat feeding in laboratory animals; therefore, high-fat diets may increase risks for developing NIDDM. RESEARCH DESIGN AND METHODS: Subjects included 544 nondiabetic women who participated in the second examination of the Kaiser Permanente Women Twins Study (1989-1990). Fasting and 2-h post 75-g glucose load insulin levels were determined. Dietary fat intake was assessed by a food frequency questionnaire. Generalized least-squares regression analyses for unpaired twin data were used to determine the relationship between dietary fat intake and insulin levels after adjustment for total calories, age, several behavioral variables, and in some models, percentage of body fat and waist-to-hip ratio. Associations of dietary fat intake with insulin levels were also evaluated within the subset of monozygotic twin pairs (n = 164 pairs) after removal of genetic influences by regression analysis of intrapair differences. RESULTS: Among the 544 individual women, a 20 g/day increase in total dietary fat was associated with a higher fasting insulin level (9% [P < 0.001] before and 6% [P < 0.01] after adjustment for the obesity variables). Higher intakes of saturated fat, oleic acid, and linoleic acid were each positively related to higher fasting insulin values. The relation of dietary fat with fasting insulin was significantly attenuated among physically active women compared with those who were sedentary (P = 0.04), even after adjustment for obesity. Only saturated fat intake was significantly associated with 2-h postglucose load insulin level before (P = 0.004) but not after adjustment for obesity. Within identical twin pairs, total dietary fat was positively related to fasting insulin before (P = 0.03) but not after adjustment for obesity (P = 0.11). CONCLUSIONS: High intake of total dietary fat is positively related to relative fasting hyperinsulinemia in nondiabetic women, particularly those who are sedentary. This effect appears to be partly mediated by the relation of dietary fat with obesity.

Postdischarge utilization of medical services by high-risk infants: experience in a large managed care organization.

BACKGROUND: Infants discharged from intensive care nurseries are a high-risk infant (HRI) population known to have increased utilization of medical services. Most studies tracking HRIs have been based on data obtained from individual chart review or direct patient contact. Given the high cost of such studies, it is desirable to develop less costly methods to track such infants. OBJECTIVES: Our goals were: (1) to identify an HRI cohort at two neonatal intensive care units; (2) to identify a control group of infants not meeting HRI criteria; and (3) to measure outpatient and inpatient utilization in both controls using computerized files in a managed care organization. METHODS: Using California Children's Services criteria as our starting point, we established an HRI definition. From a 1-year birth cohort of 7579 infants at two facilities, we identified 250 infants meeting the HRI definition at two neonatal intensive care units during 1990. We then matched the HRIs with a cohort of 896 randomly selected control newborns (those not meeting the HRI definition). Using organizational computer files and state of California death certificate tapes, we followed these infants until February 28, 1992. We measured the number of hospitalizations, total number of hospital days, and total number of outpatient visits and expressed these outcomes as rates per person-year. We also measured postdischarge mortality. RESULTS: The rate of hospitalization in the HRI group was 6.07 times (95% confidence interval [CI], 4.74-7.77) that in the control group. The utilization of hospital days by the HRI population (hospital days per 1000 person-months) was 13.24 times higher (95% CI, 11.00-16.04). The outpatient visit rate was 1.40 times higher (95% CI, 1.36-1.45) in the HRI population. CONCLUSION: Our findings in a large managed care organization corroborate previous studies showing that hospitalization rates are significantly higher among HRIs. In our study population, outpatient utilization was significantly higher as well. Our study also demonstrates the feasibility of using computerized files to study outcomes in selected pediatric populations. These methods can be used for epidemiologic studies, interventional trials, and planning for resource allocation.

Outpatient management practices associated with reduced risk of pediatric asthma hospitalization and emergency department visits.

OBJECTIVES: Effective outpatient care is believed to prevent hospitalization and emergency department (ED) visits resulting from childhood asthma. The aim of this study was to suggest priority areas for intervention by identifying outpatient management practices associated with the risk of these adverse outcomes in a large population. METHODS: This case-control study included children aged 0 to 14 years with asthma who were members of a regional health maintenance organization. Cases were children undergoing either a hospitalization or an ED visit for asthma during the study period. Control subjects were children with asthma without a hospitalization or an ED visit during the study period who were matched to patients on age, gender, and number of asthma-related hospitalizations in the past 24 months. Data on provider and parent asthma management practices were collected using chart review, closed-ended telephone interviews with parents, and computerized use databases. Multivariate analyses were conducted using conditional logistic regression models. RESULTS: Data were collected on 508 cases and 990 control subjects. A total of 43% of cases were reported by their parents to have moderately severe or severe asthma, compared with 20% of control subjects. Parents of cases with hospitalization were less likely than control subjects to have a written asthma management plan (44% vs 51%) and to report washing bedsheets in hot water at least twice a month (77% vs 86%). Cases with hospitalization were more likely to have a nebulizer (74% vs 56%). In the final multivariate model, race/ethnicity was not associated with having had either a hospitalization or an ED visit, as was lower socioeconomic status. Having a written asthma management plan [odds ratio (OR): 0.54; 95% confidence interval (CI): 0.30, 0.99] and washing bedsheets in hot water at least twice a month (OR: 0.45; 95% CI: 0.21, 0.94) were associated with reduced odds of hospitalization. Having a written asthma management plan (OR: 0.45; 95% CI: 0.27, 0.76) and starting or increasing medications at the onset of a cold or flu were associated with reduced odds of making an ED visit. CONCLUSIONS: Practices that support early intervention for asthma flare-ups by parents at home, particularly written management plans, are strongly associated with reduced risk of adverse outcomes among children with asthma.

Trends in cardiovascular disease incidence and survival in the elderly.

This study compared the age-specific incidence, postdiagnostic survival, and mortality for cardiovascular disease (CVD) in two cohorts of people aged 65 years and older. All subjects were members of a large prepaid health maintenance organization. The influence of changes in CVD risk factors on these rates also was evaluated. Trends in prevalence, incidence, postdiagnostic survival, and mortality for CVD were examined in both cohorts in 1971 and 1980. Myocardial infarction (MI), angina pectoris, stroke, and congestive heart failure (CHF) were included as CVD outcomes in this analysis. Nine-year prospective data on these diagnoses were abstracted from medical records and computerized hospitalization records for both cohorts. Age-sex-adjusted cardiovascular mortality was lower for both sexes by approximately 20% in the 1980 cohort. Overall survival did not change, whereas cancer mortality increased by 76% in women and 36% in men. With the exception of stroke, there was no increase in age-adjusted or age-specific prevalence. In men, the age-adjusted prevalence of stroke in men was 24% higher in the 1980 cohort. Age-adjusted 9-year incidence of MI, angina pectoris, stroke, and CHF did not change between cohorts in either sex Postdiagnostic, age-adjusted mortality for men with incident stroke was 24% lower in the 1980 cohort, and Postdiagnostic, age-adjusted mortality for men with incident angina was 35% lower in the 1980 cohort. Adjustment for risk factors measured at or before baseline had little influence on cohort differences in CVD incidence or duration of survival after CVD diagnosis. This study confirms other research showing a decline in CVD mortality over the past 20 years. These findings suggest that prevalent angina pectoris is increasing in men, and that survival with stroke and with angina is improving in men. Later diagnosis of incident CHF in men suggests that prevention and early detection may be postponing the development of more serious disease.

The leukocyte count: a predictor of hypertension.

In an exploratory study of 1031 persons observed to progress from normotension to essential hypertension and 1031 matched subjects who remained normotensive, the initial leukocyte count (WBC) was found to be related to the development of hypertension, with risk increased 40% (95% confidence interval 12-82%) in persons in the highest as compared to the lowest quartile of WBC. This relationship proved to be largely independent of body mass index, body fat distribution, alcohol and tobacco consumption, and parental history of hypertension. An increased WBC may reflect greater sympathetic tone or may directly increase peripheral vascular resistance by impeding circulation through small blood vessels. If confirmed, this study adds another condition to the growing list for which the WBC is predictive. This simple, cheap test should be considered for inclusion in prospective epidemiological studies of many different diseases.

The impact of aging and chronic disease on use of hospital and outpatient services in a large HMO: 1971-1991.

OBJECTIVES: To examine overall and diagnosis-specific trends in the use of inpatient and outpatient medical services (1970-1988) among older members of a large HMO. DESIGN: Two cohorts of approximately 3000 persons aged 65 or older in 1971 and 1980 were compared for hospital and outpatient utilization during 9-year follow-up periods (1971-79 and 1980-88). All subjects were evaluated for vital status throughout the follow-up period as well. PARTICIPANTS: All 6057 subjects were members of the Northern California Kaiser Permanente Medical Care Program in 1971 or 1980. The study sample was sex-age stratified (65-69,70-79,80+) at baseline. MEASUREMENTS: Data on demographics, outpatient health services utilization, categories of outpatient utilization and disease diagnoses were obtained from membership lists or medical chart review; inpatient utilization, including admitting and discharge diagnosis, length of stay, and number of hospital days was assessed from computerized hospitalization records. RESULTS: Hospital discharge rates (sex-age adjusted) increased by 12% between cohorts, with the largest increases at the oldest ages. There was a 25% increase among women and a 9% increase among men. Length of stay decreased by 20%. Hospitalization for ischemic heart disease decreased by 17%. Congestive heart failure (CHF) discharge rates (sex-age adjusted) were 92% higher in the 1980-88 cohort. For diagnoses related to nursing home institutionalization and frailty, discharge rates were significantly higher in the 1980-88 cohort: pneumonia (+34%), urinary tract infections (+104%), dehydration (+110%), osteoarthritis (+64%), syncope (+246%), leg cellulitis (+70%). In-hospital survival improved, but overall percent of readmissions also increased by 4%; readmissions for CHF increased by 13% and those for conditions of frailty by 120%. Overall outpatient visits increased by 17%. Use of laboratory tests (+57%) and outpatient surgeries (+99%) increased for all age strata in 1980-88 compared with 1971-79. CONCLUSIONS: While overall outpatient and inpatient utilization has largely decreased over the past 30 years, as a result of economic factors and improved treatments for some major diseases, there has been an increase in utilization among older people. Hospitalization for diagnoses associated with end-stage cardiovascular disease (CHF), musculoskeletal disease, frailty and iatrogenic aspects of institutionalization are clearly increasing substantially. The largest impact of aging on health care may be the result of institutionalization and its sequelae. Improved treatment for cardiovascular disease may also be leading to increased utilization at later stages in the disease process.

Statistical process control geometric Q-chart for nosocomial infection surveillance.

Several authors have proposed the use of statistical process control charting methods for the surveillance of endemic rates of nosocomial infections. The principal goal of such a charting program is to recognize any increase of the endemic rate to an epidemic rate as soon as possible after the change occurs. However, many of the statistical process control charting methods that have been proposed are based on classical charting principles that are effective largely for processes for which sufficient historical data are available. These methods require that a fairly large data set, taken while the infection rate was stable at a low endemic value, must be available to begin the charting process. These data are used both to confirm the appropriateness of the probability distribution and to make a control chart for the infection process based on the distribution. However, such data sets are often not available. The purpose of this article is to inform and demonstrate to readers that recent research in statistics has developed modern statistical process control methods that can be used effectively with or without such prior data. These methods make possible much more effective nosocomial infection surveillance programs that will give timely warnings of the onsets of epidemics or evidence of the effectiveness of infection control initiatives. These warnings will permit earlier correction initiatives and thus avoid much liability.

Reduced mortality associated with long-term postmenopausal estrogen therapy.

OBJECTIVE: To compare all-cause and specific-cause mortality rates in women who had or had not used long-term postmenopausal estrogen replacement therapy (ERT). METHODS: We identified women who used long-term postmenopausal ERT and compared them with a sample of age-matched postmenopausal nonusers. Through linking of these subjects' medical record numbers to various data bases, we examined survivorship and cause of death among estrogen users and nonusers. The risk of death in 232 postmenopausal women who began ERT within 3 years of menopause and used it for at least 5 years was compared with that of 222 age-matched postmenopausal nonusers. In the users, the mean length of estrogen use was 17.1 years. RESULTS: Statistically significant reductions in all-cause mortality were found in users compared with nonusers. For death from any cause, the age-adjusted relative risk (RR) and associated 95% confidence interval (CI) in estrogen users was 0.54 (0.38-0.76). The reduction in all-cause mortality was largely due to reductions in coronary heart disease (RR 0.40, CI 0.16-1.02) and other cardiovascular disease (RR 0.27, CI 0.10-0.71). Overall cancer mortality was similar in the two groups (RR 0.85, CI 0.46-1.58), although estrogen users had a higher risk of death from breast cancer (RR 1.89, CI 0.43-8.36) and lower risk of death from lung cancer (RR 0.22, CI 0.04-1.15). CONCLUSION: Long-term ERT use is associated with lower all-cause mortality and confers this apparent protection primarily through reduction in cardiovascular disease.

Myocardial infarction in users of low-dose oral contraceptives.

OBJECTIVE: To determine the relationship between the use of low-dose (less than 50 micrograms estrogen) oral contraceptives (OC) and myocardial infarction. METHODS: In this population-based case-control study, all incident myocardial infarctions in women, ages 15-44 years who were members of the Kaiser Permanente Medical Care Program, Northern and Southern California regions were ascertained during a 39-month period from 1991 through 1994. For each woman with myocardial infarction, up to three age- and facility-matched controls were chosen at random from female members. Information about OC use (predominantly low-dose preparations) was obtained in face-to-face interviews. RESULTS: There were 187 incident cases of myocardial infarction during 3.6 million woman-years of observation (incidence rate, 5.2 per 100,000 woman-years). The prevalence of several risk factors for myocardial infarction was lower in controls who were current users of OCs than in controls who were noncurrent (past and never) users. The odds ratio for myocardial infarction in current OC users compared with noncurrent users was 1.65 (95% confidence interval 0.45, 6.06) after adjustment for major risk factors and for race and ethnicity, corresponding to an excess risk of less than one case per 100,000 woman-years. The study had 80% power to detect a relative risk of 2.3 (one-sided test, alpha = .05). The odds ratio of myocardial infarction in past OC users was not elevated. CONCLUSION: With respect to myocardial infarction, low-dose oral contraceptives can be used safely by women who lack risk factors for coronary heart disease.

The cost of health conditions in a health maintenance organization.

In this retrospective cohort analysis of all adults who were members of Kaiser Permanente, Northern California, between July 1995 and June 1996 (N = 2,076,303), the authors estimated the prevalence, average annual costs per person, and percentage of total direct medical expenditures attributable to each of 25 chronic and acute conditions. Ordinary least squares regression was used to adjust for age, gender, and comorbidities. The costs attributable to the 25 conditions accounted for 78 percent of the health maintenance organization's total direct medical expense for this age-group. Injury accounted for a higher proportion (11.5 percent) of expenditures than any other single condition. Three cardiovascular conditions--ischemic heart disease, hypertension, and congestive heart failure--together accounted for 17 percent of direct medical expense and separately accounted for 6.8 percent, 5.7 percent, and 4.0 percent, respectively. Renal failure ($22,636), colorectal cancer ($10,506), pneumonia ($9,499), and lung cancer ($8,612) were the most expensive conditions per person per year.

Quality of primary care practice in a large HMO according to physician specialty.

OBJECTIVE: To determine whether physician specialty was associated with differences in the quality of primary care practice and patient satisfaction in a large, group model HMO. DATA SOURCES/STUDY SETTING: 10,608 patients ages 35-85 years, selected using stratified probability sampling from the primary care panels of 60 family physicians (FPs), 245 general internists (GIMs), and 55 subspecialty internists (SIMs) at 13 facilities in the Kaiser Permanente Medical Care Program of Northern California. Patients were surveyed in 1995. STUDY DESIGN: A cross-sectional patient survey measured patient reports of physician performance on primary care measures of coordination, comprehensiveness, and accessibility of care, preventive care procedures, and health promotion. Additional items measured patient satisfaction and health values and beliefs. PRINCIPAL FINDINGS: Patients were remarkably similar across physician specialty groups in their health values and beliefs, ratings of the quality of primary care, and satisfaction. Patients rated GIMs higher than FPs on coordination (adjusted mean scores 68.0 and 58.4 respectively, p<.001) and slightly higher on accessibility and prevention; GIMs were rated more highly than SIMs on comprehensiveness (adjusted mean scores 76.4 and 73.8, p<.01). There were no significant differences between specialty groups on a variety of measures of patient satisfaction. CONCLUSIONS: Few differences in the quality of primary care were observed by physician specialty in the setting of a large, well-established group model HMO. These similarities may result from the direct influence of practice setting on physician behavior and organization of care or, indirectly, through the types of physicians attracted to a well-established group model HMO. In some settings, practice organization may have more influence than physician specialty on the delivery of primary care.

Differences in resource use and costs of primary care in a large HMO according to physician specialty.

OBJECTIVE: To determine if primary care physician specialty is associated with differences in use of health services. DATA SOURCES: Automated outpatient diagnostic, utilization, and cost data on 15,223 members (35-85 years of age) of a large group model HMO. STUDY DESIGN: One-year prospective comparison of primary care provided by 245 general internists (GIMs), 60 family physicians (FPs), and 55 subspecialty internists (SIMs) with case-mix assessed during a nine-month baseline period using Ambulatory Diagnostic Groups. PRINCIPAL FINDINGS: Adjusting for demographics and case mix, patients of GIMs and FPs had similar hospitalization and ambulatory visit rates, and similar laboratory and radiology costs. Patients of FPs made fewer visits to dermatology, psychiatry, and gynecology (combined visit rate ratio: 0.86, 95% CI: 0.74-0.96). However, they made more urgent care visits (rate ratio 1.19, 95% CI: 1.07-1.23). Patients of SIMs had higher hospitalization rates than those of GIMs (rate ratio 1.33, 95% CI: 1.06-1.68), greater use of urgent care (rate ratio: 1.14, 95% CI: 1.04-1.25), and higher costs for pharmacy (cost ratio: 1.17, 95% CI: 0.93-1.18) and radiologic services (cost ratio: 1.14, 95% CI: 1.01-1.30). The hospitalization difference was due partly to the inclusion of patients with specialty-related diagnoses in panels of SIMs. Radiology and pharmacy differences persisted after excluding these patients. CONCLUSIONS: In this uniform practice environment, specialty differences in primary care practice were small. Subspecialists used slightly more resources than generalists. The broader practice style of FPs may have created access problems for their patients.