PubMed Health⌕ Search

Biomedical subjects

R B Keller

Publications and source records attributed to R B Keller.

At least 19 recordsLinked to original sources
PubMed1998

Prevalence and predictors of long-term work disability due to carpal tunnel syndrome.

The objective of this study was to describe patterns and predictors of work absence in the prospective, community-based Maine Carpal Tunnel Study. Three hundred fifteen patients with carpal tunnel syndrome (CTS) were recruited from physicians' offices throughout Maine. The patients completed questionnaires at entry and after 6, 18, and 30 months. The questionnaires included scales measuring symptom severity, functional status, general and mental health status, exposure to physical stressors, work status, and other indicators. The analyses examined univariate and multivariate correlates of work absence. The mean age was 43, 72% of subjects were female, 71% underwent carpal tunnel release, and 45% were receiving Workers' Compensation. Fifty-two percent worked in managerial or technical occupations, 15% in service occupations, and 13% in heavy labor or machine operation. Forty-five percent of patients changed jobs or were absent from work (aside from postoperative recovery) during the 30-month follow-up. In multivariate logistic regression models, correlates of work absence at 18 months included worse functional status of the hand at study entry and at 6-month follow-up, involvement of an attorney at the time of enrollment (P < 0.002 for each), and work absence at 6 months (P = 0.03). Worse upper extremity functional status and having a contested Workers' Compensation claim are critical predictors of work absence and should be principal targets of interventions to reduce work disability in CTS.

Absenteeism↗
PubMed1998

Maine Carpal Tunnel Study: small area variations.

As is the case for all elective procedures, small area variations occur in rates of surgery for carpal tunnel syndrome. A 1993 analysis of Maine data demonstrates that carpal tunnel release rates across population-based service areas varied 3.5-fold, from 0.82 to 2.87 per thousand. Four areas had rates significantly higher and 2 were significantly lower than the state average of 1.44 per thousand. Among many potential factors influencing variations, physician practice patterns appear to be the major contributor.

Carpal Tunnel Syndrome↗
PubMed1998

Maine Carpal Tunnel Study: outcomes of operative and nonoperative therapy for carpal tunnel syndrome in a community-based cohort.

A prospective, community-based, observational study of the outcome of surgical and nonoperative management was conducted. The study included 429 patients with carpal tunnel syndrome recruited in physicians' offices throughout Maine. Patients were assessed at baseline and at 6, 18, and 30 months following presentation using validated scales that measured symptom severity, functional status, and satisfaction. Seventy-seven percent of eligible survivors from the original cohort were monitored for 30 months. Surgically treated patients demonstrated improvements of 1.2 to 1.6 points on the 5-point Symptom Severity and Functional Status scales (23% to 45% improvement in scores), which persisted over the 30-month follow-up period. The nonoperatively managed patients showed little change in clinical status at 6, 18, and 30 months. While workers' compensation recipients had worse outcomes than nonrecipients, 36 of 68 (53%) workers' compensation recipients were completely or very satisfied with the results of the procedure 30 months after surgery. There were no significant differences in outcome between patients treated with endoscopic versus open carpal tunnel release. Among worker's compensation recipients, 12 of 68 (18%) surgical patients and 4 of 32 (13%) nonoperatively treated patients remained out of work because of carpal tunnel syndrome at 30 months. Thus, carpal tunnel surgery offered excellent symptom relief and functional improvement in this prospective community-based sample, irrespective of the surgical approach, even in workers' compensation recipients. Work absence remained high in both surgically and nonoperatively managed workers' compensation recipients.

Aged↗
PubMed1998

Physician partnering in Maine: an update from the Maine Medical Assessment Foundation.

BACKGROUND: The Maine Medical Assessment Foundation (MMAF) has involved the participation of hundreds of physicians in study groups to analyze data on small-area variation and assess physician decision-making patterns. In 1991 the MMAF model was replicated across a tri-state area (Maine, New Hampshire, and Vermont) in an effort called the Outcomes Dissemination Project. THE OUTCOMES DISSEMINATION PROJECT: Five specialty study groups, each meeting three times a year, examined local and national utilization data and guidelines and research findings, participated in outcomes studies and patient education, and disseminated their findings through society presentations and other feedback efforts. Physician surveys indicated that all but one of the study groups were successful in making their existence and activities known to the broader medical community. PARTNERING WITH PURCHASERS AND MEDICAID: In the "Partnership Projects," the MMAF and relevant study groups, collaborating with a Maine business coalition, analyzed variations in utilization across employers, types of health plans (for example, health maintenance organization, preferred provider organization, fee-for-service), and small areas. Through this process, coalition members learned about small-area analysis and the implications of variations in utilization for cost and quality. The first year of the collaborative projects focused on several issues, including the development and dissemination of a practice guideline for the care of patients with chest pain; the dissemination of a practice guideline for the treatment of patients with acute low back pain; and a project designed to address cesarean section rates. SUMMARY: Innovative partnerships have been crafted to allow the tenets of the MMAF's working relationships with the medical community to remain intact, ensuring the ongoing interest and cooperation of hundreds of Maine physicians.

Community Networks↗
PubMed1998

Comparative responsiveness of generic versus disease-specific and weighted versus unweighted health status measures in carpal tunnel syndrome.

OBJECTIVES: The authors evaluated the relative responsiveness to change of generic versus disease-specific and unweighted versus weighted health status measures in carpal tunnel syndrome (CTS). METHODS: Data were obtained from 196 subjects followed in a prospective community-based cohort study in Maine who underwent carpal tunnel release (The Maine Carpal Tunnel Syndrome Study). Patients were evaluated before and 6 months after surgery. The disease-specific, unweighted severity score was derived from the validated Carpal Tunnel Syndrome Assessment Questionnaire. Patients were asked to rate the importance of each symptom included in the severity score. Each severity question was weighted by its importance, creating a disease-specific weighted score. Generic instruments were the SF-36, SF-12, and a Quality of Life Rating Scale. Sensitivity to change was calculated with the standardized response mean (SRM, mean change/standard deviation of change) as well as the effect size (ES, mean change/standard deviation of baseline values). The ability of the instruments to distinguish clinically important differences was assessed by correlating the changes in scores with global ratings on satisfaction and perceived improvement as external criteria. RESULTS: The disease-specific weighted score (SRM: 1.56, ES: 1.99) was more responsive than the unweighted score (SRM: 1.36, ES: 1.57). The Quality of Life Rating Scale, SF-36, and SF-12 subscales were less sensitive to change, with standardized response means and effect sizes that ranged from -0.23 to 0.88. The ability to distinguish clinically important differences was higher for the two disease-specific scales. The coefficients of correlation with the external criteria ranged from 0.50 to 0.56 for the unweighted score and 0.56 to 0.62 for the weighted score and were significantly stronger than the correlations between external measures and the most responsive subscale of the SF-36 (Bodily Pain subscale, r = 0.36). The SF-12 health survey performed as well as the SF-36 in term of responsiveness and ability to distinguish clinically important change. CONCLUSIONS: Disease-specific measures were superior to generic measures in capturing clinical change after carpal tunnel release, and a weighted score was slightly more responsive than the unweighted score. The SF-12 showed comparable psychometric properties compared with the longer 36-item Short-Form Survey.

Adult↗
PubMed1997

Predictors of return to work following carpal tunnel release.

Little is known about factors that predict return to work following carpal tunnel release. Patients enrolled in a prospective, community-based study of carpal tunnel syndrome in Maine were evaluated with standardized questionnaires preoperatively and 6 months following carpal tunnel release. Univariate and multivariate analyses were performed to identify baseline factors associated with work disability 6 months following surgery. Thirty-one of 135 patients (23%) were out of work because of CTS 6 months following surgery. The predominant preoperative variables associated with work absence due to CTS 6 months postoperatively in logistic regression analyses were Workers' Compensation, work absence preoperatively, and worse mental health status (p < or = 0.01 for each). In analyses that considered postoperative as well as preoperative variables, persistence of symptoms following surgery was the most striking predictor of failure to return to work due to CTS (p < 0.0001). Preoperative correlates of less complete relief of symptoms in multivariate models included involvement of an attorney, milder preoperative symptom severity, preoperative work absence (p < 0.005 for each) and exposure to hand intensive work (p = 0.04). These data indicate that economic and psychosocial variables have a strong influence upon both return to work and the extent of symptom relief 6 months following surgery for carpal tunnel syndrome.

Adult↗
PubMed1997

Do physicians do what they say? The inclination to test and its association with coronary angiography rates.

OBJECTIVE: Efforts to evaluate variations in cardiac procedures have focused on patient factors and differences in health care delivery systems. We wanted to assess how physicians' inclination to test patients with coronary artery disease influences utilization patterns. SETTING AND SUBJECTS: Physicians and the populations of Maine, New Hampshire, and Vermont. DESIGN: We conducted a survey of 263 family practitioners, internists, and cardiologists residing in 57 hospital service areas in Maine, New Hampshire, and Vermont. Using patient scenarios, we assessed the clinicians' inclinations to test during the evaluation of patients with coronary artery disease. Self-reported testing intensities were used to create three indices: a Catheterization Index, an Imaging Exercise Tolerance Test (ETT) Index, and Nonimaging ETT Index. Using administrative data, age- and gender-adjusted population-based coronary angiography rates were calculated. Physicians were assigned to low (2.9/1,000), average (4.2/1,000), and high (5.8/1,000) coronary angiography rate areas, based on where they practice. Analysis of variance techniques were used to assess the relation of the index scores to the population-based coronary angiography rates and to physician specialties. RESULTS: There was a positive relationship between the population-based coronary angiography rates and the self-reported scores of the Catheterization Index (p < .005) and the Imaging ETT Index (p = .01), but none was found for the Non-imaging ETT Index (p = .10). These relationships were evident in subanalyses of cardiologists and internists, but not of family practitioners. CONCLUSIONS: Self-reported testing intensity by physicians is related to the population-based rates of coronary angiography. This relationship cuts across specialties, suggesting that there is a "medical signature" for the evaluation of patients with coronary artery disease.

Adult↗
PubMed1997

Patients' preferences and their relationship with satisfaction following carpal tunnel release.

Patients' preferences for specific health outcomes in carpal tunnel syndrome (CTS) and their association with demographic factors and satisfaction with the results of surgery after 6 months were evaluated. Two hundred fifty subjects with CTS and enrolled in a prospective community-based cohort study in Maine completed a preference questionnaire before surgery. Patients were asked to specify the single most important reason they decided to undergo surgery and to rate the importance of improvement in 10 areas, including lessening of specific symptoms and improvement in specific functional states related to CTS. The single most important reason for CTS patients to have surgery was relief of night pain (37%), followed by relief of numbness (21%) and relief of daytime pain (13%). When patients were asked to rate the importance of obtaining relief from specific symptoms and improvement in specific functional states, relief of numbness received the highest rating, with 94% of the patients answering that it was extremely or very important. Workers' compensation recipients, patients with less than a college level of education, and patients with more severe symptoms and functional impairment at baseline assigned higher importance to symptom relief and functional improvement. Controlling for other predictors, higher preference for improved strength was associated with lower satisfaction with the results of the surgery at 6 months. Most CTS patients undergoing surgery have realistic preferences for health outcomes that are influenced by demographic and clinical characteristics; however, physicians should pay attention to unrealistic preferences that might influence patients' satisfaction with surgical results.

Carpal Tunnel Syndrome↗
PubMed1997

Prognostic value of a hand symptom diagram in surgery for carpal tunnel syndrome.

OBJECTIVE: To evaluate symptom patterns on a hand diagram as predictors of surgical outcome in carpal tunnel syndrome (CTS). METHODS: 202 patients with CTS enrolled in a prospective, community based cohort study in Maine completed a hand symptom diagram before surgery and at 6 month followup. They were asked to mark on the hand diagram the location of 3 symptoms: pain, numbness/tingling (NT), and "other" symptoms. The diagram was first divided into 6 regions following a standardized procedure. For the 6 regions, symptom patterns were identified separately for each of the 3 symptoms. Outcomes 6 months after surgery were expressed as the percentage of change on the Symptom Severity Scale and Function Status Scale of the Carpal Tunnel Syndrome Assessment Questionnaire, and the satisfaction with the results of the surgery. RESULTS: Several distinct symptom patterns were associated with the 3 principal outcomes in univariate and multivariate analysis. In linear regression models controlling for the baseline severity of symptoms and function, as well as other predictors, the hand symptom pattern variables accounted for 30, 14, and 24%, respectively, of the total explained variance in satisfaction, symptom severity, and functional status. Patients receiving Workers' Compensation (37% of the cohort) had more wrist pain and NT of the arm, and less pain involving the arm and upper palm. This group also had worse outcomes and were less satisfied with surgery. Drawing expansion was associated with a low score on the SF-36 mental health subscale. However, psychological impairment was not associated with a worse outcome. CONCLUSION: Symptom patterns identified preoperatively with a hand symptom diagram help to predict the outcome of carpal tunnel release.

Adult↗
PubMed1996

The Quebec Task Force classification for Spinal Disorders and the severity, treatment, and outcomes of sciatica and lumbar spinal stenosis.

STUDY DESIGN: A prospective cohort study of patients in Maine with sciatica and lumbar spinal stenosis treated surgically and nonsurgically. SUMMARY OF BACKGROUND DATA: In 1987, the Quebec Task Force on Spinal Disorders proposed a diagnostic classification to help make clinical decisions, evaluate quality of care, assess prognosis, and conduct research. OBJECTIVES: To assess the Quebec Task Force classification's ability to stratify patients according to severity and treatment at baseline, and to assess changes over time in health-related quality of life, including symptoms, functional status, and disability. METHODS: Five hundred sixteen patients participating in the Maine Lumbar Spine Study who completed baseline and 1-year follow-up evaluations were classified successfully according to the Quebec Task Force classification. Patient characteristics and treatments were compared across Quebec Task Force classification categories. Changes in health-related quality of life over 1 year were assessed according to Quebec Task Force classification category and type of treatment. RESULTS: Among patients with sciatica (n = 370), higher Quebec Task Force classification categories (from 2, pain radiating to the proximal extremity, to 6, sciatica with evidence of nerve root compression) were associated with increased severity of symptoms at baseline. There was no association between Quebec Task Force classification and baseline functional status. Quebec Task Force classification was associated strongly with the likelihood of receiving surgical treatment (P < or = 0.005). Among patients with sciatica treated nonsurgically, improvement at 1 year in back-specific and generic physical function increased with higher Quebec Task Force classification category (P < or = 0.05). Only a nonsignificant trend was observed for surgically treated patients. Patients with lumbar spinal stenosis (Quebec Task Force classification 7, n = 131) had baseline features and outcomes distinct from patients with sciatica. CONCLUSIONS: For patients with sciatica, the Quebec Task Force classification was highly associated with the severity of symptoms and the probability of subsequent surgical treatment. Nonsurgically treated patients in Quebec Task Force classification categories reflecting nerve root compression had greater improvement than those with pain symptoms alone. Among surgical patients, the Quebec Task Force classification was not associated with outcome. These results provide validation for the classification and its wider adoption. Nonetheless, improved diagnostic classifications are needed to predict outcomes better in patients with sciatica who undergo surgery.

Adolescent↗
PubMed1996

The Maine Lumbar Spine Study, Part I. Background and concepts.

STUDY DESIGN: This paper describes the background and factors that led to the development and implementation of the Maine Lumbar Spine Study, a prospective cohort study of patients undergoing surgical and nonsurgical treatment of herniated lumbar disc with sciatica and symptomatic spinal stenosis. OBJECTIVES: To define the factors leading to the study and the methods of designing and implementing a community-based effectiveness study to evaluate the outcomes of herniated lumbar intervertebral disc and spinal stenosis. SUMMARY OF BACKGROUND DATA: Variations in the utilization of surgery for these conditions and physicians' uncertainty regarding the best way to manage them resulted in support of a community-based study of the effectiveness of treatment alternatives. METHODS: A prospective cohort design was used. Methods of patient enrollment, data collection, management, and analysis are described. An innovative method of ascertaining the representativeness of the enrolled versus nonenrolled patient population is presented. RESULTS: The importance of developing community-based networks of physicians is discussed. CONCLUSIONS: These networks play an important role in analyzing practice pattern variations and in stimulating and participating in effectiveness research. Because effectiveness studies must be conducted at the community level, mechanisms must be developed with which to support and implement these efforts.

Cohort Studies↗
PubMed1996

The Maine Lumbar Spine Study, Part II. 1-year outcomes of surgical and nonsurgical management of sciatica.

STUDY DESIGN: The Maine Lumbar Spine Study is a prospective cohort study of patients recruited from the practices of orthopedic surgeons, neurosurgeons, and occupational medicine physicians throughout Maine. OBJECTIVE: To assess 1-year outcomes of patients with sciatica believed to be due to a herniated lumbar disc treated surgically or nonsurgically. SUMMARY OF BACKGROUND DATA: Lumbar spine surgery rates vary by geographic region and may reflect uncertainty about optimal clinical use. METHODS: Eligible consenting patients participated in a baseline interview performed by study personnel and then were mailed follow-up questionnaires at 3, 6, and 12 months. Clinical data were obtained from a physician questionnaire. Outcomes included patient-reported symptoms of leg and back pain, functional status, disability, quality of life, and satisfaction with care. RESULTS: Five hundred seven patients with sciatica, 275 treated surgically and 232 treated nonsurgically initially, were enrolled. Surgically treated patients, on average, had more severe symptoms and had more severe physical and imaging findings than nonsurgically treated patients at entry. Although few surgically treated patients had mild symptoms and few nonsurgically treated patients had severe symptoms, about half in each treatment group had symptoms that fell into a moderate category. At the 1-year evaluation, improvement in symptoms, functional status, and disability were found in both treatment groups. However, surgically treated patients reported significantly greater improvement. For the predominant symptom, either back or leg pain, 71% of surgically treated and 43% of nonsurgically treated patients reported definite improvement (P < 0.001). This effect was even greater after adjustment for differences between treatment groups at entry (relative odds of definite improvement, 4.3; P < 0.001). For patients with moderate symptoms and abnormal physical examination findings, surgical treatment also resulted in greater improvement than nonsurgical treatment. However, there was little difference in the employment or workers' compensation status of patients treated surgically versus nonsurgically (5% vs. 7% unemployed at 1-year follow-up if employed at entry [P = 0.68]; 46% vs. 55% receiving workers' compensation at 1-year follow-up if receiving it at entry [P = 0.30] for surgical and nonsurgical management, respectively). For patients with mild symptoms, the benefits of surgical and nonsurgical treatment were similar. CONCLUSIONS: Although surgically treated patients were on average more symptomatic at entry, there was substantial overlap in symptoms between surgically treated and nonsurgically treated patients. Surgically treated patients with sciatica reported substantially greater improvement at 1-year follow-up. However, employment and compensation outcomes were similar between the two treatment groups, and surgery appeared to provide little advantage for the subset of patients with mild symptoms. These results should be interpreted cautiously, because surgical treatment was not assigned randomly. Long-term follow-up will determine if these differences persist.

Adolescent↗
PubMed1996

The Maine Lumbar Spine Study, Part III. 1-year outcomes of surgical and nonsurgical management of lumbar spinal stenosis.

STUDY DESIGN: A prospective cohort study of patients with lumbar spinal stenosis recruited from the practices of orthopedic surgeons and neurosurgeons throughout Maine. OBJECTIVE: To assess 1-year outcomes of patients with lumbar spinal stenosis treated surgically or nonsurgically. SUMMARY OF BACKGROUND DATA: No randomized trials and few nonexperimental studies have compared surgical and nonsurgical treatment of patients with lumbar spinal stenosis. The authors' goal was to assess 1-year outcomes of patients with lumbar spinal stenosis treated surgically or nonsurgically. METHODS: Eligible, consenting patients participated in baseline interviews and were then mailed follow-up questionnaires at 3, 6, and 12 months. Clinical data were obtained from a physician questionnaire. Outcomes included patient-reported symptoms of leg and back pain, functional status, disability, and satisfaction with care. RESULTS: One hundred forty-eight patients with lumbar spinal stenosis were enrolled, of whom 81 were treated surgically and 67 treated nonsurgically. On average, patients in the surgical group had more severe imaging findings and symptoms and worse functional status than patients in the nonsurgical group at entry. Few patients with mild symptoms were treated surgically, and few patients with severe symptoms were treated nonsurgically. However, of the patients with moderate symptoms, a similar percent were treated surgically or nonsurgically. One year after study entry, 28% of nonsurgically and 55% of surgically treated patients reported definite improvement in their predominant symptoms (P = 0.003). For patients with moderate symptoms, outcomes for surgically treated patients were also improved compared with those of nonsurgically treated patients. Surgical treatment remained a significant determinant of 1-year outcome, even after adjustment for differences between treatment groups at entry (P = 0.05). The maximal benefit of surgery was observed by the time of the first follow-up evaluation, which was at 3 months. Although few nonsurgically treated patients experienced a worsening of their condition, there was little improvement in symptoms and functional status compared with study entry. CONCLUSIONS: At a 1-year evaluation of patient-reported outcomes, patients with severe lumbar spinal stenosis who were treated surgically had greater improvement than patients treated nonsurgically. Comparisons of outcomes by treatment received must be made cautiously because of differences in baseline characteristics. A determination of whether the outcomes observed persist requires long-term follow-up.

Adult↗
PubMed1996

The association between local diagnostic testing intensity and invasive cardiac procedures.

OBJECTIVE: To determine the extent to which geographic variation in invasive cardiac procedures can be explained by the variable use of diagnostic testing. DESIGN: A population-based cohort study using Medicare Part B data (physician services). SETTING AND SUBJECTS: Procedure data for all Medicare beneficiaries in northern New England. MAIN OUTCOME MEASURES: Twelve coronary angiography service areas were constructed for Medicare beneficiaries in northern New England. Age- and sex-adjusted utilization rates were developed for three procedure categories: total stress test, coronary angiography, and revascularization. Total stress tests were further stratified into nonimaging and imaging procedures (eg, thallium). Tests performed in follow-up to invasive procedures were excluded (eg, stress test following revascularizations). Linear regression was used to assess the relationship between procedure categories. RESULTS: A tight positive relationship was found between total stress test rates and the rates of subsequent coronary angiography (R2=0.61, P<.005). Most of the variance was explained by imaging stress tests (R2=0.50, P<.02). A strong relationship was found between coronary angiography and revascularization (R2=0.82, P<.001). Finally, a clear relationship between total stress tests and subsequent revascularizations was also found (R2=0.55, P<.006). CONCLUSION: The population-based rates of diagnostic testing largely explained the variance associated with subsequent therapeutic interventions. Our results suggest that local testing intensity is an important determinant of the variable use of invasive cardiac procedures.

Cardiovascular Diseases↗
PubMed1996

Application of outcomes research in occupational low back pain: the Maine Lumbar Spine Study.

Outcomes research represents an expansion of traditional clinical research to include issues of cost and quality of care in usual clinical practice, emphasizing outcomes that matter most to patients. In low back disorders, outcomes research has focused on the lack of reliable information to support much of clinical practice and has demonstrated marked variability in the treatment of these common problems. The Maine Lumbar Spine Study represents an example of an outcomes research study to investigate the treatment of patients with sciatica in usual clinical practice. Because low back symptoms are a frequent cause of occupational disability, Workers' Compensation patients were explicitly oversampled. Baseline features were significantly different in those patients who were receiving Workers' Compensation versus those who were not. Efforts to compare outcomes by disability status need to control for these differences. Whereas most Workers' Compensation patients were still receiving disability compensation regardless of treatment at 6 months, patients who were treated surgically were more likely to have come off disability and returned to work than nonsurgically treated patients. Long-term follow-up is necessary to determine whether these differences persist.

Adult↗
PubMed1996

Measuring outcomes.

Explore the source record for details and available documents.

Health Status Indicators↗
PubMed1996

Basic epidemiology of fractures of the upper and lower limb among Americans over 65 years of age.

Current knowledge regarding the basic epidemiology of fractures is largely limited to a few fracture sites, notably those of the hip and distal forearm. To clarify the patterns of incidence of limb fractures in the elderly, we used data from a 5% sample of the U.S. Medicare population over age 65 years during the years 1986-1990. We identified incident fractures of the proximal humerus, other parts of the humerus, proximal radius/ ulna, shaft of the radius/ulna, distal radius/ulna, pelvis, hip, other parts of the femur, patella, ankle, and other parts of the tibia/fibula from diagnoses and procedures coded on claims for inpatient services, outpatient facility use, and physician services. We used Poisson regression to investigate the relation between demographic factors and fracture risk at these sites. Fractures at the hip were the most common, accounting for 38% of the fractures identified. The proximal humerus, distal radius/ulna, and ankle also were common fracture sites. A pattern of rapidly rising rates with age was seen for fractures of the pelvis, hip, and other parts of the femur among women. Fractures distal to the elbow or knee, however, had, at most, modest increases in incidence with age over 65 years. For each of the fractures studied, women had higher rates than men of the same race, and whites generally had higher rates than blacks of the same gender. Gender-related differences in risk were larger among whites than among blacks, and racial differences in risk were more marked among women than among men.

Aged↗