Evaluation of the effect of performance monitoring and feedback on care process, utilization, and outcome.

OBJECTIVE: We evaluated a program of performance measurement and monitoring by assessing care process, utilization of services, and outcomes. RESEARCH DESIGN AND METHODS: Information on 63,264 diabetic individuals who were continuously enrolled as members of Kaiser Permanente Southern California from 1 January 1994 to 31 December 1997 was used to evaluate the program. Time trends in testing for glycemic test and control and screening for dyslipidemia, use of lipid-lowering drugs, and microalbuminuria were evaluated as measures of care process. Time trends in hospitalization, outpatient appointments, prescriptions, and laboratory tests were evaluated as measures of utilization. Outcomes were hospitalization for myocardial infarction, ischemic stroke, and lower-limb amputation. RESULTS: Between 1994 and 1997, improvements were evident in the process measures. The mean number of hospitalizations and the mean and median number of outpatients visits did not change. The mean number of laboratory tests increased from 13.2 in 1994 to 23.6 in 1997. The mean number of prescriptions for any medication increased from 19.7 to 24.3. Hospitalization rates for myocardial infarction did not change, but rates increased for ischemic stroke and lower-limb amputation. CONCLUSIONS: Our findings suggest that measurement and monitoring of clinical performance can bring about modest improvements in measures of the processes of care in the absence of financial incentives, centrally driven interventions, and specialty care for all patients. In our setting, process improvements were associated with higher utilization of laboratory services and more prescriptions without an immediate return in terms of lower hospital utilization.

Duration of warfarin anticoagulant therapy and the probabilities of recurrent thromboembolism and hemorrhage.

To determine the association of duration of warfarin anticoagulant therapy after hospitalization for venous thromboembolism with the probability of recurrent thromboembolism and with the risks of hemorrhage, medical records of 2,422 patients hospitalized in 1970 through 1980 with pulmonary embolism, thrombophlebitis, or both were reviewed. Multivariate life-table analyses were performed for 370 patients who had positive results of venography or pulmonary angiography, or who had lung scanning evidence of a "high probability" of pulmonary embolism and no history of the disease. For these patients, warfarin therapy for more than six weeks was not associated with a lower risk of recurrent thromboembolism when compared with warfarin therapy for one through six weeks (risk of recurrence for seven to 26 weeks of treatment 0.8; 95 percent confidence limits 0.3 and 2.5; risk of recurrence for more than 26 weeks of treatment 1.1; 95 percent confidence limits 0.4 and 3.1). The longer the warfarin therapy, the higher the risk of medically important complications from therapy. From one week through five years, the probability of major hemorrhage increased almost linearly: 10 percent for 12 weeks, 18 percent at one year, 26 percent at two years, and 41 percent at five years. This study suggests that intensive, long-term warfarin anticoagulation, in patients with a first episode of venous thromboembolism and no predisposing condition, is associated with more toxicity than efficacy and should be abandoned.

Meta-analysis and endocrinology.

It is likely that more studies that use meta-analysis will be published in the endocrinologic literature. Major strengths of meta-analysis are the systematic ascertainment of research on a given topic and the explicit delineation of reasons for accepting or rejecting studies as a basis for drawing conclusions. The tendency of meta-analysis to focus on a single estimate of effect and to ignore heterogeneity are problems both with the conduct of meta-analysis and the way in which it is interpreted. Meta-analysis cannot overcome bias in the original studies. It is difficult to perform a good meta-analysis and easy to perform a bad one. The critical reader should not be overawed by the results of a meta-analysis. Reading a meta-analysis should not substitute for careful reading of the primary studies on which the meta-analysis is based. Meta-analysis should not be used to stifle the conduct of original research.

The influence of intrapartum antibiotics on the clinical spectrum of early-onset group B streptococcal infection in term infants.

OBJECTIVE: The use of intrapartum antibiotics to prevent early-onset group B streptococcal (EOGBS) infection has left pediatricians in a quandary about the appropriate evaluation and treatment of infants at risk for this infection. The aim of this study was to determine whether intrapartum antibiotic prophylaxis changed the constellation and timing of onset of clinical signs of group B streptococcal (GBS) infection in term infants. METHODOLOGY: We conducted a retrospective chart review of infants who had EOGBS infection and were born in Southern California Kaiser Permanente Hospitals from 1988 through 1996. Objective criteria were used to ascertain maternal risk of infection, intrapartum antibiotic prophylaxis, and onset of clinical signs of infection. RESULTS: Three hundred nineteen infants with EOGBS sepsis, bacteremia, or clinically suspected infection were identified from a population of 277 912 live births. Of the 172 term infants with culture-positive infection who had clinical signs of infection, 95% exhibited them in the first 24 hours of life. All of the infants exposed to intrapartum antibiotics became ill within the first 24 hours of life. CONCLUSIONS: Exposure to antibiotics during labor did not change the clinical spectrum of disease or the onset of clinical signs of infection within 24 hours of birth for term infants with EOGBS infection. A 48-hour stay is not required to monitor asymptomatic term infants exposed to intrapartum antibiotics for onset of GBS infection.

Malignant hypertension in women aged 15 to 44 years and its relation to cigarette smoking and oral contraceptives.

To study the factors associated with malignant hypertension in young women, we reviewed the medical records of all women aged 15 to 44 years who were hospitalized in any of 15 affiliated hospitals from 1971 to 1980 with malignant hypertension. There were 15 new patients in this period, an incidence of 0.5 per 100,000 women of these ages per year. Four women had underlying renal disease. All of the remaining 11 used oral contraceptives, smoked cigarettes, or both. Nine were white, and 6 women had normal blood pressure in the 24 months before hospitalization. These findings support a relation of oral contraceptive use and cigarette smoking to an increased risk of malignant hypertension in young women.

PIP: To study the factors associated with malignant hypertension in young women, we reviewed the medical records of all women ages 15-44 who were hospitalized in any of 15 affiliated hospitals from 1971-80 with malignant hypertension. There were 15 new patients in this period, an incidence of 0.5/100,000 women of these ages/year. 4 women had underlying renal disease. All of the remaining 11 used oral contraceptives (OCs), smoked cigarettes, or both. 9 were white and 6 women had normal blood pressure in the 24 months prior to hospitalization. These findings support a relation of OC use and cigarette smoking to an increased risk of malignant hypertension in young women.

Coronary heart disease and estrogen replacement therapy. Can compliance bias explain the results of observational studies?

The overall risk/benefit of estrogen replacement therapy (ERT) is strongly dependent on assumptions about the effect of ERT on coronary heart disease (CHD). The belief that ERT causes a substantial reduction in the risk of CHD is widespread. The studies that provide support for this belief are all nonexperimental ones. Three analyses of data from two randomized clinical trials of drug treatments for CHD have examined the association of compliance with total mortality in persons who complied with the taking of placebo. In these analyses, compliance with the taking of a placebo was associated with a reduction in overall mortality of the same magnitude as the reduction in the risk of CHD in users of ERT. The benefit of compliance with placebo was not reduced by adjustment for a large number of variables, both medical and sociodemographic, that might affect mortality. Users of ERT are compliers, and the possibility that compliance bias may account for some of the apparent benefit of ERT for CHD must be taken seriously.

Prothrombin time ratio and other factors associated with bleeding in patients treated with warfarin.

We previously showed that the risk of major hemorrhage in patients with venous thromboembolism treated with warfarin was strongly related to duration of anticoagulant therapy. We here report the results of a more detailed analysis of factors other than duration of warfarin therapy associated with the risk of hemorrhage in these patients. Almost 7% of patients had a major hemorrhage on warfarin and an additional 23.7% had at least one minor bleeding episode. Age, female sex, and congestive heart failure were associated with small increases in the risk of major hemorrhage but not with the risk of minor bleeding. A prothrombin time ratio greater than 2.5 was associated with a fourteen-fold increase in the risk of a major hemorrhage (95% CI 5.1, 42.7), but major hemorrhages occurred in patients on warfarin at all measured values of the prothrombin time ratio. Taken together with the findings from our previous analysis, the study suggests that prevention of bleeding in patients on warfarin would best be accomplished by minimizing the duration of warfarin therapy, by scrupulous monitoring of the prothrombin time ratio, and by considering the "therapeutic range" for the prothrombin time ratio to be somewhat less than 2.0-2.5.

Evidence for compensation in smokers of low yield cigarettes.

To determine the relation of smoking low yield cigarettes (tar yield less than 15.0 mg and nicotine yield less than 1.0 mg per cigarette) to the number of cigarettes smoked per day, we analysed information provided on self-administered questionnaires by 7706 current, regular cigarette smokers. The mean age at starting to smoke and the mean number of years of smoking were not consistently different in smokers of low yield compared with smokers of high yield cigarettes. In contrast, at all ages in both men and women, smokers of low yield cigarettes smoked significantly more cigarettes per day than smokers of high yield cigarettes. The differences in the number of cigarettes smoked per day between smokers of low and high yield cigarettes were small--about 3 cigarettes per day in men and about 1 1/2 cigarettes per day in women. However, these small differences might translate to the smoking of as many as one billion more packs of cigarettes per year in the United States alone. The potential beneficial effect of the smoking of cigarettes whose yield per cigarette is lower must be weighed against a possible adverse effect of the smoking of more cigarettes if the relation between smoking low yield cigarettes and smoking more cigarettes is causal.

Noncontraceptive estrogens and mortality: long-term follow-up of women in the Walnut Creek Study.

The effect of postmenopausal estrogen use on mortality is an important and controversial subject. To address it, we analyzed data from a ten- to 13-year mortality follow-up of the 3437 women enrolled in the Walnut Creek Contraceptive Drug Study who never used either estrogens or oral contraceptives and the 2656 women who used estrogens but not oral contraceptives. By 1983, 109 estrogen users and 110 nonusers had died. After adjusting for age, mortality in estrogen users (1.64 per 1000 woman-years) was lower than in nonusers (2.06 per 1000 woman-years) for all categories of cause of death except cancer. The lower mortality from accidents, suicide, and homicide in estrogen users has no plausible biologic explanation, and the observation suggests that life-style differences between estrogen users and nonusers account at least in part for their lower mortality. On the other hand, the observation that the relative risk of mortality due to cardiovascular disease was 0.5 in estrogen users, after adjustment for age and other cardiovascular disease risk factors, is suggestive.

Morbidity among breech infants according to method of delivery.

To determine whether vaginally born breech infants are at increased risk for morbid events as compared with breech infants delivered by cesarean, we studied 1240 singleton breech infants without congenital anomalies delivered in Northern California Kaiser Permanente Medical Care Program hospitals during 1976-1977. Medical record review provided information on indications for method of delivery, delivery complications and injuries, neonatal complications, and neurologic sequelae up to 4 years of age. The relative risk estimates for asphyxia (1.0; 95% confidence interval 0.7, 1.4), head trauma (1.6; 95% confidence interval 0.2, 17.0), neonatal seizures (0.8; 95% confidence interval 0.1, 7.1), cerebral palsy (1.6; 95% confidence interval 0.2, 17.4), and developmental delay (2.0; 95% confidence interval 0.9, 4.4) for vaginally born compared with cesarean-delivered infants indicated that vaginally born infants were not at increased risk for these outcomes. We used multiple logistic regression to control for confounding variables. The adjusted relative risk estimate for the combined-outcome category of head trauma, neonatal seizures, cerebral palsy, mental retardation, or spasticity was 0.5 in vaginally delivered infants (95% confidence interval 0.1, 3.2). When all morbid outcomes were considered in combination, the adjusted relative risk estimate was 0.9 for vaginally delivered infants (95% confidence interval 0.6, 1.4).

In-hospital maternal mortality in the United States: time trends and relation to method of delivery.

To study time trends in maternal mortality in the United States and to attempt to compare the risk of cesarean with vaginal delivery, information from the Professional Activities Study of the Commission on Professional and Hospital Activities for 3 years-1970, 1974, and 1978-was reviewed. For all deliveries, mortality per 100,000 deliveries declined from 25.7 in 1970 to 14.3 in 1978. For vaginal deliveries, mortality per 100,000 deliveries declined from 20.4 to 9.8. For cesarean deliveries, mortality per 100,000 births decreased more than for vaginal deliveries, from 113.8 to 40.9. Mortality for deliveries with no mention of complications, lacerations, or uterine rupture declined significantly from 1970 to 1978. Mortality for deliveries complicated by dystocia or malpresentation declined significantly from 1970 to 1974, but failed to decline thereafter. Mortality for deliveries complicated by a previous cesarean or by antepartum hemorrhage did not decline significantly from 1970 to 1978. For all complications with a sufficient number of vaginal and cesarean deliveries, except deliveries complicated by malpresentation or antepartum hemorrhage, mortality was at least twice as high in cesarean as in vaginal deliveries. Based on a comparison of mortality after a previous cesarean with mortality for all vaginal deliveries with no complication, the authors conclude that cesarean delivery is probably neither less than 2 nor more than 4 times more hazardous than vaginal delivery.

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.

Steroid hormone contraception and bone mineral density: a cross-sectional study in an international population. The WHO Study of Hormonal Contraception and Bone Health.

OBJECTIVE: To evaluate relationships between bone mineral density and use of steroid hormonal contraceptives. METHODS: This was a multicenter cross-sectional study in seven centers in three regions of the developing world from April 1994 to June 1997. Women 30-34 years old attending family planning clinics, with at least 24 months of lifetime use of combined oral contraceptives (OC), depot-medroxyprogesterone acetate (DMPA), or levonorgestrel implants, or no or only short-term (less than 6 months) use of steroid hormonal contraceptives, had bone mineral density (BMD) measured at the distal radius and the midshaft of the ulna using single-photon x-ray absorptiometry. RESULTS: In the study, 2474 women were examined. For OC use, adjusted mean BMD was significantly higher in short-term, current users compared with women who never used hormonal contraceptives. For DMPA and levonorgestrel implants, adjusted mean BMD was statistically significantly lower in short-term current users compared with those who never used hormonal contraceptives. For all three hormonal methods, there were no significant differences in BMD between past users of hormonal contraceptives and never users, even among those who had used the methods for 4 or more years. The magnitude of changes in BMD was small and less than one standard deviation (SD) from the mean of those who never used steroid contraceptives. CONCLUSION: This study suggests that hormonal contraceptive use by young adult women is associated with small changes in BMD that occur early after initiation of use and are reversible.

Changes in providers' views and practices about emergency contraception with education.

OBJECTIVE: To assess changes in the prescribing practices, knowledge, attitudes, and perceptions of health care providers after an educational program about emergency contraception. METHODS: Health care providers completed self-administered questionnaires before and 1 year after full implementation of the project. The 102 providers who completed both questionnaires were physicians (64%) and mid-level professionals from 13 San Diego County Kaiser Permanente medical offices working in departments such as obstetrics and gynecology, primary care, and emergency medicine. RESULTS: The frequency of prescription for emergency contraceptive pills increased significantly from baseline to follow-up. There was an increase of almost 20% in the percentage who prescribed emergency contraception at least once a year. Knowledge also improved significantly, and perceptions of barriers to prescribing emergency contraceptive pills within the health maintenance organization decreased significantly. In contrast, attitudes about emergency contraception showed little change. CONCLUSION: This study suggests that providers who participate in in-service training and other aspects of a demonstration project show changes in perceptions, knowledge, and behavior. However, findings also suggest that significant gaps remain in knowledge about medications, side effects, and mode of action. It is likely that many providers in other health care settings also need additional information and training concerning protocols of emergency contraception provision and its modes of action and effects.

Obesity and cholecystectomy among women: implications for prevention.

Almost all studies of the association of body mass with the risk of gallbladder disease have found that risk increases with obesity. Some studies report a nonlinear relationship of body mass to gallbladder disease risk, and some report that the association of excess body mass with increased risk of gallbladder disease is confined to younger women. We examined the association of body mass with the risk of cholecystectomy based on data from 14 years of follow-up of 16,638 women. The cholecystectomy rate was much greater in the highest quintile of Quetelet's body mass index than in other quintiles. The striking association of high rates of cholecystectomy with the highest quintile of Quetelet's body mass index was observed in each age group studied. The low incidence of cholecystectomy in young, thin women made the relative risk of cholecystectomy in women in the fifth quintile of body mass highest in women under 25 years of age. The observation that cholecystectomy rates are not linearly associated with obesity suggests that women in the highest quintiles of body mass could be considered a reasonable group in which strategies for prevention of gallbladder disease might be tested.

PIP: The Walnut Creek Contraceptive Drug Study which began in 1968 and ended in 1972, monitored 16,638 women as part of a multiphasic health checkup and collected information on oral contraceptive (OC) use; smoking, alcohol use; other habits, and family and medical histories. The number of women years was determined in each age and weight category. The rates of gallbladder disease were similarly calculated. The Cox proportional hazards model was utilized as implemented in the multivariate life table analyses. In addition, Quetelet's body mass index as a measure of obesity was applied. 16,240 women had an intact gallbladder; of these 432 were hospitalized with diagnosis of gallbladder disease. The diagnosis was confirmed in 423 women: 98% had cholelithiasis accompanied by pathologic evidence of cholecystitis. The gallbladder disease rate increased with age, with body mass index, and it was dramatically higher in women in the highest quintile of Quetelet's body mass index than in other quintiles at all ages. A nonlinear association of cholecystectomy with obesity persisted. At least 500,000 cholecystectomies are carried out in the US annually, and the risk is much higher in women in the uppermost quintile, thus it is reasonable to suggest that preventive efforts should focus on this group.

Physiologic measures in men with and without vasectomies.

PIP: Blood pressure, hematologic and blood chemistry measurements, and serologic tests for syphilis were performed on 4385 vasectomized and 13,155 age- and race-matched, nonvasectomized men who had multiphasic health checkups as members of a large prepaid medical care program in California. The study was carried out from 1977 through 1980. Age distributions of the vasectomized and nonvasectomized men were identical, 33% of both groups being 40 years old or younger, 33% being 41-50, and 34% being 51 or older. 37% of the vasectomized men had their vasectomies 10 or more years before their examination. Systolic blood pressure was slightly but significantly lower and potassium was slightly but significantly higher in the vasectomized group. All other differences were small and not biologically important. These data agree with other studies of vasectomized humans, in which no association of vasectomy with adverse health outcomes, including atherosclerosis, has been found.^ieng

Employment and low birth weight in black women.

Recent studies have not shown adverse effects of employment during pregnancy on the pregnancy outcome, but most have not successfully distinguished factors due to work-related physical and mental strain from differences in social, economic and cultural characteristics between employed and non-employed women. We analyzed data from the Alameda County Case-Control Study of Low Birth Weight to address the question of employment as a risk factor for low birth weight in more depth. Subjects in the study were singleton infants without congenital malformations born in 1987 to black or white, non-hispanic residents of Alameda County, a large urban community in California. Cases were all infants weighing less than 2500 g at birth; controls were chosen at random from the infants weighing 3000 g or more. The mothers of 377 black cases, 389 black controls, 233 white cases and 239 white controls could be located and interviewed. This analysis excluded the white women because evidence of response bias in regard to employment was found for them but not for black women. The percentage of black women in the study who were employed during pregnancy was 49.9%. Employment was associated with a significantly lower relative risk of low birth weight before and after controlling for age, parity, smoking, heavy alcohol use, prior low birth weight infants and low pre-pregnancy weight. After adjustment for known risk factors for low birth weight, employment explained more of the variance of low birth weight than other commonly used measures of 'social class', including income, education and marital status.(ABSTRACT TRUNCATED AT 250 WORDS)

Worldwide variations in the lifetime probability of reproductive cancer in women: implications of best-case, worst-case, and likely-case assumptions about the effect of oral contraceptive use.

Cancer incidence in countries representative of three patterns of reproductive cancer and age-specific mortality was used to estimate the effect of oral contraceptive use on the lifetime probability of reproductive cancer under three sets of assumptions about the effects of oral contraceptives. Under the set of assumptions considered likely, oral contraceptives were estimated to reduce or increase only slightly the lifetime probability of any reproductive cancer in each setting. Under worst-case assumptions, oral contraceptives were estimated to increase the lifetime probability of reproductive cancer only modestly in settings with low cancer rates and in settings with high rates of breast, ovarian, and endometrial cancer, but it might have a large impact on lifetime probability of reproductive cancer in settings with high cervical cancer rates. Under best-case assumptions, oral contraceptives were estimated to decrease the lifetime probability of reproductive cancer in each setting; this reduction was estimated to be greatest in settings where endometrial and ovarian cancer incidence are high.

PIP: Researchers applied published data on cancer incidence and age specific mortality to standard life table techniques to estimate the lifetime probability of developing reproductive cancer for women living in countries representative of 3 patters of risk of reproductive cancer and for long term oral contraceptives (OC) users under best case, worst case, and likely case assumptions. The reproductive cancers included breast, ovarian, endometrial, and cervical cancers. The data consisted of urban women from China, Japan, United States (California), England, Wales, Costa Rica, and Colombia. Under the likely case assumption, OCs just barely reduced or increased the lifetime probability of any reproductive cancer in any setting. Further, under the worst case scenario, OCs increased the lifetime probability or reproductive cancer moderately in countries with low cancer rates (Asian countries) and in countries with high rates of breast, ovarian, and endometrial cancer (Western Europe, North America, and Australia). Yet in countries with high cervical rates (South and Central America), OC use significantly affected the lifetime probability of reproductive cancer. The best case scenario revealed that OCs decreased lifetime probability of reproductive cancer in each country, especially those countries where endometrial and ovarian cancer incidences were great. The analysis also showed that OC use has the greatest effect on lifetime probability of reproductive cancer, be it positive or negative, in countries with high underlying rates of reproductive cancer. Further it demonstrated that the effect of OC use will most likely be small in countries with low incidence of reproductive cancers. Overall the researchers felt reassured about OC use and reproductive cancer. Even though long term OC use increases the risk of breast cancer in young ages.

Oral contraceptive use and myocardial infarction.

Epidemiologic studies of current oral contraceptive (OC) use and myocardial infarction (MI) have been contradictory and confusing. This comprehensive review of the epidemiologic literature attempts to draw conclusions about the risk of myocardial infarction in oral contraceptive users, focusing on recent, methodologically rigorous studies of the topic. Recent studies are consistent in showing a higher relative risk of MI in current OC users who smoke compared with current OC users who do not smoke. Studies in aggregate suggest that the relative risk of MI is higher in current OC users with hypertension than in current OC users without hypertension. Recent studies do not show a relationship between higher estrogen dose and higher risk of MI, but the effect of estrogen dose and progestogen type and dose are difficult to separate. The limited data on the risk of MI in current users of low estrogen OC do not allow a firm conclusion about the possibility that progestogen type might affect the risk of MI in current users. Past OC use does not increase or decrease the risk of MI.

PIP: Epidemiologic studies of the association between oral contraceptive (OC) use and the risk of myocardial infarction (MI) have produced conflicting results. Estimates of MI risk in current OC users have ranged from 0.87 to 5.01. However, many of these studies include serious methodologic flaws, including a restriction to idiopathic cases, failure to assess potential confounders such as smoking, and lack of attention to the problem of effect modification. This article is based on a comprehensive search of all studies published in 1960-97 that estimated the relative risk of MI or acute coronary artery disease in relation to past or current OC use and considered the effects of OC type, smoking, diabetes, a family history of MI, hypercholesterolemia, hypertension, age, and other risk factors. Studies that reported modification of the effect of current OC use by smoking consistently show higher relative risks than those that found no evidence of effect modification by smoking. Overall, the data suggest that the risk of MI for current OC use is probably increased by a factor of 1.2-1.8 in women who do not have high blood pressure and do not smoke. Current OC use in both smokers and women with high blood pressure is associated with a 5- to 10-fold increase in MI use compared to that of nonsmokers or nonhypertensive women who do not use OCs. There is no evidence of differences in the relative risk of MI between the different progestogens contained in OCs with less than 50 mcg of estrogen. Any conclusions about the relationship between OC formulation and MI risk remain tentative, however, as a result of small sample sizes, low power, inconsistency of results, and problems separating the effect of estrogen and progestogen type and dose.

Norplant implants and cardiovascular disease.

A 1995 publication on serious adverse events in users of Norplant implants submitted to the Food and Drug Administration's MedWatch Spontaneous Reporting System reported 14 hospitalizations for stroke in Norplant users. This number was higher than expected. This is a report on the association of current use of Norplant implants with stroke and myocardial infarction (MI) based on a pooled analysis of data from two population-based, case-control studies conducted in the US. All data collection for these two studies occurred after approval of Norplant implants for marketing in the US in December 1990. The methods of the individual studies are detailed in prior publications.

PIP: A 1995 publication submitted to the US Food and Drug Administration's MedWatch Spontaneous Reporting System in 1995 identified 14 hospitalizations for stroke in US Norplant contraceptive implant users. This paper reports the findings of a pooled analysis of data from two large population-based case-control studies conducted in California and Washington. Of the 518 stroke patients and their 1547 healthy controls, only 1 stroke patient, 1 ischemic stroke patient, and 3 controls were current Norplant users. After adjustment for age, the odds ratio (OR) for stroke in current compared with noncurrent users of Norplant was 1.0 (95% confidence interval (CI), 0.1-9.2). In addition, 307 myocardial infarction patients and their 1048 controls were available for analysis. Of these, 1 case and 1 control were current users of Norplant. The age-adjusted OR for myocardial infarction in current compared with noncurrent Norplant users was 3.5 (95% CI, 0.2-56.5). The low prevalence of Norplant use in these studies, combined with the rarity of cardiovascular events in women of reproductive age, limited the statistical power of the pooled analysis to determine whether Norplant use increases, decreases, or has no effect on the risk of cardiovascular disease.