A cost-utility analysis of atezolizumab in the second-line treatment of patients with metastatic bladder cancer.

Background: Despite initial promising results, the IMvigor211 clinical trial failed to demonstrate an overall survival (os) benefit for atezolizumab compared with chemotherapy as second-line treatment for metastatic bladder cancer (mbc). However, given lessened adverse events (aes) and preserved quality of life (qol) with atezolizumab, there might still be investment value. To evaluate that potential value, we conducted a cost-utility analysis (cua) of atezolizumab compared with chemotherapy from the perspective of the Canadian health care payer. Methods: A partitioned survival model was used to evaluate atezolizumab compared with chemotherapy over a lifetime horizon (5 years). The base-case analysis was conducted for the intention-to-treat (itt) population, with additional scenario analyses for subgroups by IMvigor-defined PD-L1 status. Health outcomes were evaluated through life-year gains and quality-adjusted life-years (qalys). Cost estimates in 2018 Canadian dollars for systemic treatment, aes, and end-of-life care were incorporated. The incremental cost-effectiveness ratio (icer) was used to compare treatment strategies. Parameter and model uncertainty were assessed through sensitivity and scenario analyses. Per Canadian guidelines, cost and effectiveness were discounted at 1.5%. Results: For the itt population, the expected qalys for atezolizumab and chemotherapy were 0.75 and 0.56, with expected costs of $90,290 and $8,466 respectively. The resultant icer for atezolizumab compared with chemotherapy was $430,652 per qaly. Scenario analysis of patients with PD-L1 expression levels of 5% or greater led to a lower icer ($334,387 per qaly). Scenario analysis of observed compared with expected benefits demonstrated a higher icer, with a shorter time horizon ($928,950 per qaly). Conclusions: Despite lessened aes and preserved qol, atezolizumab is not considered cost-effective for the second-line treatment of mbc.

Annual relative increased in inpatient mortality from antimicrobial resistant nosocomial infections in Thailand.

Antimicrobial resistance is a major health threat worldwide as it brings about poorer treatment outcome and places economic burden to the society. This study aims to estimate the annual relative increased in inpatient mortality from antimicrobial resistant (AMR) nosocomial infections (NI) in Thailand. A retrospective cohort study was conducted at Ramathibodi Hospital, Bangkok, Thailand, over 2008-2012. Survival model was used to estimate the hazard ratio of mortality of AMR relative to those patients without resistance (non-AMR) after controlling for nine potential confounders. The majority of NI (73.80%) were caused by AMR bacteria over the study period. Patients in the AMR and non-AMR groups had similar baseline clinical characteristics. Relative to patients in the non-AMR group, the expected hazard ratios of mortality for patients in the AMR group with Acinetobacter baumannii, Escherichia coli, Pseudomonas aeruginosa and Staphylococcus aureus were 1.92 (95% CI 0.10-35.52), 1.25 (95% CI 0.08-20.29), 1.60 (95% CI 0.13-19.10) and 1.84 (95% CI 0.04-95.58), respectively. In the complete absence of AMR bacteria, this study estimated that annually, in Thailand, there would be 111 295 fewer AMR cases and 48 258 fewer deaths.

Economic evaluation of a publicly funded hepatitis A travel vaccination program in Ontario, Canada.

BACKGROUND: Hepatitis A virus (HAV) causes acute liver infection and is spread through the fecal-oral route. Travel to countries in HAV-endemic regions (e.g., Asia and Latin America) is a well-described risk factor for infection. Currently, Ontario publicly funds hepatitis A vaccination for some populations at high risk of HAV infection but not for all travellers to endemic countries. The objective of this study was to determine the cost-effectiveness of expanding publicly funded HAV vaccination to people planning travel to HAV-endemic regions, from the Ontario healthcare payer perspective. METHODS: We conducted a cost-utility analysis comparing an expanded high-risk publicly-funded hepatitis A vaccination program including funded vaccine for travellers to endemic regions to the current high risk program in Ontario. A Markov state transition model was developed, including six possible health states. Model parameters were informed through targeted literature searches and included hepatitis A disease probabilities, utilities associated with health states, health system expenditures, and vaccine costs. Future costs and health outcomes were discounted at 1.5%. Primary outcomes included cost, incremental cost-effectiveness ratio (ICER) and quality adjusted life years (QALYs) over a lifetime time horizon. We conducted one-way, two-way, and probabilistic sensitivity analysis. RESULTS: The expanded high risk HAV vaccine program provided few incremental health gains in the travel population (mean 0.000037 QALYs/person), at an incremental cost of $124.31. The ICER of the expanded program compared to status quo is $3,391,504/QALY gained. The conclusion of the model was robust to changes in key parameters across reasonable ranges. CONCLUSIONS: The expanded vaccination program substantially exceeds commonly accepted cost-effectiveness thresholds. Further research concerning possible cost-effective implementation of high-risk travel hepatitis A vaccination should focus on a more integrated understanding of the risk of acquiring hepatitis A during travel to endemic regions (e.g., purpose, length of stay).

Policy brief: How can the settings used to provide care to older people be balanced?

The optimal balance between institutional, home-based and community care for older adults requires an effective mix of organizational, funding and delivery mechanisms for target populations. This spans health and social care, and the coordination of care must respect older people's care preferences and that of their families and friends as well as limits on the available resources to support and fund service provision. Care settings used to provide long-term care for older people and how they are defined vary greatly across Europe. This policy brief addresses the appropriate balance between three main components of long-term care: home care services; institutional care (formal and informal sectors); and care provided by family and friends (informal care). The dramatic upward trend in the cost and use of long-term care, the projected impact of ageing populations and the prevalence of age-related chronic disease and dependency ratios have catalysed proposals to redesign the funding, organization and delivery of affordable, effective and equitable health and social care for older people.

The economic burden of disabling hip and knee osteoarthritis (OA) from the perspective of individuals living with this condition.

OBJECTIVE: To estimate the direct and indirect arthritis-attributable costs to individuals with disabling hip and/or knee osteoarthritis (OA). METHODS: An established population cohort with disabling hip and/or knee OA from two regions of Ontario, Canada was surveyed to determine participant and caregiver costs related to OA, and the predictors of these costs. RESULTS: The response rate was 87.2%. Of 1378 respondents, 1258 had OA (mean age 73.1 yr, range 59-100). Sixty per cent (n = 758) reported OA-related costs. Among these individuals, the average annual cost was 12,200 dollars(CDN dollars in 2002, where 1.00 CDN dollar approximately 0.81 US dollar). Time lost from employment and leisure by participants and their unpaid caregivers accounted for 80% of the total. Men were less likely than women to report costs (adjusted odds ratio 0.54, P < 0.0001), but when they did their expenditures were significantly higher (P = 0.004). Greater disability was associated with higher costs: compared with individuals with WOMAC total scores <15, those with scores > or = 55 were 15 times more likely to report costs, and their costs were 3 times greater (both P < 0.0001). Both the young (<65 yr) and very old were more likely to incur costs (P < 0.0001), and when they did their costs were higher (P < 0.001). CONCLUSION: Costs incurred were mainly for time lost from employment and leisure, and for unpaid informal caregivers. Failure to value such indirect costs significantly underestimates the true burden of OA. Costs increased with worsening health status and greater OA severity. After adjustment, men were less likely to incur costs, possibly due to greater social resources.

Registered nurses' experiences with an evidence-based home care pathway for myocardial infarction clients.

OBJECTIVES: To obtain home health nurses' comments on an evidence-based care pathway for post myocardial infarction. DESIGN: A qualitative design was used. SETTING: Culturally diverse, lower income area of a large city. PARTICIPANTS: All home health nurses from one nursing agency who participated in a comparative study on the impact of the evidence-based care pathway. RESULTS: The largest number of comments made by the nurses were related to the beneficial impact of the pathway on the provision of quality nursing care and on increased job satisfaction. The home health nurses reported that the pathway increased clients' knowledge of medications and diet. In addition, they commented that they were able to use the pathway effectively because of the training they received from the inpatient cardiac nurses. CONCLUSIONS: This qualitative study demonstrates the benefits of investing in the implementation of best practice guidelines by home health nurses. However, nursing associations, such as the Canadian Community Health Nurses Initiatives Group, will need to continue to champion for additional funds to support the additional expenses incurred.

Socio-economic status and surgery in children: myringotomies and tonsillectomies in Ontario, Canada, 1996-2000.

AIM: To examine the relation between socio-economic status and (1) receipt of paediatric otolaryngological surgery, and (2) inclusion of adjuvant procedures. METHODS: Using data on myringotomies with insertion of tympanostomy tube and tonsillectomies for all children in Ontario, Canada, from 1996 to 2000, and census data on socio-economic status, we examined the association between socio-economic status and (1) the probability of surgery (myringotomy or tonsillectomy), and (2) the probability that surgery was accompanied by an adjuvant procedure. RESULTS: Lower socio-economic status was associated with increased likelihood that a child's initial surgery was a tonsillectomy rather than a myringotomy (odds ratio per unit increase in the deprivation index = 1.09, p = 0.01, confidence interval 1.06-1.11), and with increased likelihood that those children having a myringotomy would undergo a tonsillectomy during the same hospitalization (odds ratio 1.14, p < 0.0001, confidence interval 1.11-1.16). Children from neighbourhoods with larger immigrant populations were less likely to receive either procedure (odds ratios per 1% increase in the proportion of immigrants = 0.97 (p < 0.0001, confidence interval 0.96-0.97) for myringotomies and 0.97 (p < 0.0001, confidence interval 0.97-0.98) for tonsillectomies). CONCLUSIONS: Socio-economic status was associated with treatment selection for the two most common paediatric surgical procedures. Further research should examine whether differences in treatment arise at the level of the primary care physician, the specialist, and/or are due to parental preference.

Physician and population determinants of rates of middle-ear surgery in Ontario.

CONTEXT: Small-area variations in surgical rates raise concerns about access to care, treatment appropriateness, and the quality and cost of care. OBJECTIVE: To measure small-area variations in rates of myringotomy with insertion of tympanostomy tubes (TTs) and to identify determinants of rate variation. DESIGN AND SETTING: Retrospective analyses using hospital discharge data for patients who had undergone a myringotomy with insertion of TT by county in Ontario between April 1, 1996, and March 31, 1999. Information on possible determinants was taken from a survey of otolaryngologists and primary care physicians in 1996 and from the 1996 Canadian census and physician demographic databases for 1996-1999. PARTICIPANTS: A total of 75 358 hospitalizations for TT placement of children and adolescents (aged

Incremental cost analysis of ambulatory clinic and home-based intravenous therapy for patients with multiple myeloma.

BACKGROUND: Patients with multiple myeloma and other forms of cancer receiving pamidronate via intravenous (IV) infusion at the Hamilton Regional Cancer Centre in Hamilton, Ontario, Canada face 2 treatment options: they can have their entire treatment completed at the clinic using traditional IV therapy (e.g. IV bag and pole) or they can have the treatment initiated at the clinic and then return home to complete the treatment utilising a portable and disposable IV therapy device. OBJECTIVE: To perform a cost analysis of these 2 treatment options. PERSPECTIVE: Societal. METHODS AND PATIENTS: Data on all patients with multiple myeloma who attended the Hamilton Regional Cancer Centre for pamidronate therapy from November 1, 1997 to October 31, 1998 were collected from clinic records. As almost all of these patients with multiple myeloma completed their IV therapy at home, comparison to clinic-based therapy was based on derived cost estimates. Cost data, where possible, were acquired from the Hamilton Regional Cancer Centre's records. A sensitivity analysis was also conducted. RESULTS: In the base-case scenario for the study period, the incremental cost of the infusion device and training in Canadian dollars ($Can; 1998 values) for the 48 patients (299 cycles) who had their infusion initiated at the clinic but completed at home was $Can 5.50/cycle ($Can 4,636 for the 299 cycles). If these 48 patients had had their entire infusion at the clinic, the incremental costs of overtime treatment, parking, clinic overheads and lost work or leisure time would have been $Can 68.49/cycle ($Can 20,477 for the 299 cycles). Therefore, shifting treatment from the clinic to the home resulted in net cost savings to society of $Can 52.98/cycle ($Can 15,841 for the 299 cycles). Sensitivity analysis of best- and worst-cost scenarios did not alter the substantive findings although the relative difference between treatment options varied. In the best-case scenario, home treatment was $Can 95.97/cycle ($Can 28,696 for the 299 cycles) less costly than clinic treatment, while in the worst-case scenario, home treatment was $Can 17.19/cycle ($Can 5,141 for the 299 cycles) less costly than clinic treatment. The results also demonstrated that clinic overheads, the cost of a portable and disposable infusion device and the cost of lost work and leisure time had the greatest impact on incremental costs for each treatment option. CONCLUSION: Subject to study limitations, a significant cost advantage was demonstrated through the home-based treatment option for patients with multiple myeloma. Key issues that must be addressed in future evaluations include the precise determination of clinic overheads, the valuation of lost work and/or leisure time and the direct cost of portable and disposable infusion devices.

Prospective study of the patient-level cost of asthma care in children.

Our objective was to assess the cost of asthma care at the patient level in children from the perspectives of society, the Ontario Ministry of Health, and the patient. In this longitudinal evaluation, health service use data and costs were collected during telephone interviews at 1, 3, and 6 months with parents of 339 Ontario children with asthma. Direct costs were respiratory-related visits to healthcare providers, emergency rooms, hospital admissions, pulmonary function tests, prescription medications, devices, and out-of-pocket expenses. Indirect costs were parents' absences from work/usual activities and travel and waiting time. Hospital admissions accounted for 43%, medications for 31%, and parent productivity losses for 12% of total costs from a societal perspective. Statistically significant predictors of higher total costs were worse symptoms, younger age group, and season of participation. Adjusted annual societal costs per patient in 1995 Canadian dollars varied from $1,122 in children aged 4-14 years to $1,386 in children under 4 years of age. From the Ministry of Health perspective, adjusted annual costs per patient were $663 in children over 4 years and $904 in younger children. Adjusted annual costs from the patient perspective were $132 in children over 4 years and $129 in children under 4 years. The rising incidence of pediatric asthma demands that greater attention be paid to the delivery of optimal care to this segment of the population. Appropriate methods must be used to analyze healthcare costs and the use of services in the midst of widespread healthcare reform. The quality of clinical and health policy decision-making may be enhanced by cost-of-illness estimates that are comprehensive, precise, and expressed from multiple perspectives.

The role of adjuvant adenoidectomy and tonsillectomy in the outcome of the insertion of tympanostomy tubes.

BACKGROUND: Otitis media is the most common medical problem in young children. The usual surgical treatment is myringotomy with insertion of tympanostomy tubes. There is debate about the usefulness of concomitant adenoidectomy or adenotonsillectomy. We examined the effects of these adjuvant procedures on the rates of reinsertion of tympanostomy tubes and rehospitalization for conditions related to otitis media. METHODS: Using hospital discharge records for the period 1995 through 1997, we examined the results of surgery for all 37,316 children (defined as persons 19 years of age or younger) in Ontario, Canada, who received tympanostomy tubes as their first surgical treatment for otitis media. We determined the time to the first readmission for conditions related to otitis media and the time to the first reinsertion of tympanostomy tubes. RESULTS: As compared with treatment involving the insertion of tympanostomy tubes alone, adjuvant adenoidectomy was associated with a reduction in the likelihood of reinsertion of tympanostomy tubes (relative risk, 0.5; 95 percent confidence interval, 0.5 to 0.6; P<0.001) and the likelihood of readmission for conditions related to otitis media (relative risk, 0.5; 95 percent confidence interval, 0.5 to 0.6; P<0.001). The risk of these outcomes was further reduced if an adjuvant adenotonsillectomy was performed. The effect was age-related. Children as young as one year appeared to benefit from adjuvant adenotonsillectomy; the benefit of an adjuvant adenoidectomy was apparent in two-year-olds and was greatest for children three years of age or older. CONCLUSIONS: Performing an adenoidectomy at the time of the initial insertion of tympanostomy tubes substantially reduces the likelihood of additional hospitalizations and operations related to otitis media among children two years of age or older.

Determining the need for hip and knee arthroplasty: the role of clinical severity and patients' preferences.

BACKGROUND: Area variation in the use of surgical interventions such as arthroplasty is viewed as concerning and inappropriate. OBJECTIVES: To determine whether area arthroplasty rates reflect patient-related demand factors, we estimated the need for and the willingness to undergo arthroplasty in a high- and a low-use area of Ontario, Canada. RESEARCH DESIGN: Population-based mail and telephone survey. SUBJECTS: All adults aged > or =55 years in a high (n = 21,925) and low (n = 26,293) arthroplasty use area. MEASURES: We determined arthritis severity and comorbidity with questionnaires, established the presence of arthritis with examination and radiographs, and evaluated willingness to have arthroplasty with interviews. Potential arthroplasty need was defined as severe arthritis, no absolute contraindication for surgery, and evidence of arthritis on examination and radiographs. Estimates of need were then adjusted for patients' willingness to undergo arthroplasty. RESULTS: Response rates were 72.0% for questionnaires and interviews. The potential need for arthroplasty was 36.3/1,000 respondents in the high-rate area compared with 28.5/1,000 in the low-rate area (P <0.0001). Among individuals with potential need, only 14.9% in the high-rate area and 8.5% in the low-rate area were definitely willing to undergo arthroplasty (P = 0.03), yielding adjusted estimates of need of 5.4/1,000 and 2.4/1,000 in the high- and low-rate areas, respectively. CONCLUSIONS: Demonstrable need and willingness were greater in the high-rate area, suggesting these factors explain in part the observed geographic rate variations for this procedure. Among those with severe arthritis, no more than 15% were definitely willing to undergo arthroplasty, emphasizing the importance of considering both patients' preferences and surgical indications when evaluating need and appropriateness of rates for surgery.

Changing rates of cancer. The hazards of using a single definition.

Changing cancer rates, abstracted from tumor registries, are used to make inferences about the effect of carcinogens and cancer treatments on a population-wide basis. We compared the annual age-standardized incidence rates of extremity soft tissue sarcomas from two large tumor registries using different case definitions. We identified all limb soft tissue sarcoma cases diagnosed 1973-1993 in the Ontario Cancer Registry (OCR) and the Surveillance, Epidemiology, and End Results (SEER) databases. Two case definitions for limb soft tissue sarcoma were used based on missing data, incomplete diagnostic methods and ICD-9 codes; an upper limit estimate of the rates which included all possible cases of limb soft tissue sarcoma and a lower limit estimate of the rates which included all definite cases of limb soft tissue sarcoma (with the true rates lying in between). The upper limit OCR rates showed a statistically significant decreasing linear trend (slope = -0.021, P < 0.01). Whereas the slope of the OCR lower limit regression line showed a statistically significant increase in rates (slope = 0.01, P = 0.04). Neither the upper or lower limit SEER rates had a statistically significant linear trend (slope = 0.002, P = 0.60 and slope = 0.001, P = 0.18, respectively). Case definition affects incidence rates and changing rates of cancer. Thus the use of a single case definition along with changing coding practices may alone explain changing cancer rates.

Home care in Canada: passing the buck.

This paper provides an overview of Canadian home-care utilization, highlights the health-policy assumptions that have resulted in an increasing reliance on in-home services, and assesses the current roles of the private and public sectors in the financing of home care. Significant interprovincial variations in per capita home-care expenditures and potential inequalities in access to home care call for resolution by federal and provincial governments. There is a need for consensus with respect to medically and socially necessary services that are subject to national standards, irrespective of the setting in which services are sought, received, and delivered. The development and enforcement of national home-care standards that complement the principles of the Canada Health Act would be a useful first step in ensuring that the Canadian health-care system is ready to confront the challenges of the new millennium.

Referral of children with otitis media. Do family physicians and pediatricians agree?

OBJECTIVE: To determine factors influencing family physicians' and pediatricians' decisions to refer children with recurrent acute otitis media (RAOM) and otitis media with effusion (OME) to otolaryngologists for an opinion about tympanostomy tube insertion. DESIGN: Mailed survey. SETTING: Physicians' practices in Ontario. PARTICIPANTS: Random sample of 1459 family physicians and all 775 pediatricians in the province. MAIN OUTCOME MEASURES: Physicians' reports of the influence of 17 factors on decisions to refer (more likely, no influence, less likely to refer) and number of episodes of otitis media, months with effusion, level of hearing loss, or months of continuous antibiotics without improvement prompting referral. RESULTS: Physicians agreed (> 80% concordance) on six out of 17 factors as indications for referring children with RAOM or OME. Opinions about the importance of other factors varied widely. Family physicians would refer children with otitis media after fewer episodes of illness, fewer months of effusion, lower levels of hearing loss, and fewer months of prophylactic antibiotic therapy than pediatricians (all P < .001). Pediatricians would prescribe continuous antibiotics longer (11.8 weeks) than family physicians (8.9 weeks, P < .0001), which correlated with lower referral thresholds for family physicians. CONCLUSION: Family physicians' and pediatricians' self-reported referral practices for surgical opinions on children with otitis media varied considerably. These observations raise questions about the consistency of care for children with otitis media and whether revised clinical guidelines would be helpful.

Socio-economic status and the utilisation of physicians' services: results from the Canadian National Population Health Survey.

This paper assesses the extent to which Canada's universal health care system has eliminated socio-economic barriers in the use of physician services by examining the role of socio-economic status in the differential use of specific, publicly-insured, primary and specialist care services. Data from the 1994 National Population Health Survey, a nationally representative survey, were analysed using multiple logistic regression. In order to control for the association between primary and specialist utilisation, a two-staged least squares method was used for models explaining specialist utilisation. Health need, as measured by perceived health status and number of health problems, was found to be consistently associated with increased physician utilisation, for both primary and specialist visits. Whereas the likelihood of an individual making at least one visit to a primary care physician was found to be independent of income, those with lower incomes were more likely to be frequent users of primary care, that is, make at least six visits to a primary care physician. Even after adjusting for the greater utilisation of primary care services by those in lower socio-economic groups, and, therefore, their higher exposure to the risk of referral, the utilisation of specialist visits was greater for those in higher socio-economic groups. Canadians lacking a regular medical doctor were less likely to receive primary and specialist care, even after adjustments for socio-economic variables such as income and education. Although financial barriers may not directly impede access to health care services in Canada, differential use of physician services with respect to socio-economic status persists. After adjusting for differences in health need, Canadians with lower incomes and fewer years of schooling visit specialists at a lower rate than those with moderate or high incomes and higher levels of education attained despite the existence of universal health care.

Otolaryngologists' perceptions of the indications for tympanostomy tube insertion in children.

BACKGROUND: Bilateral myringotomy with insertion of tympanostomy tubes is the most common operation that children in Canada undergo. Area variations in surgical rates for this procedure have raised questions about indications used to decide about surgery. The objective of this study was to describe the factors that influence otolaryngologists to recommend tympanostomy tube insertion in children with otitis media and their level of agreement about indications for surgery. METHODS: A survey was sent to all 227 otolaryngologists in Ontario in the fall of 1996. The influence of 17 clinical and social factors on recommendations to insert tympanostomy tubes were assessed. Case vignettes were used to determine the effect of multiple factors in decisions about the need for surgical management. RESULTS: Surveys were returned by 138 (68.3%) of the 202 eligible otolaryngologists. There was agreement (more than 90% of respondents) about 6 indications for surgery: persistent effusion, a lack of improvement after 3 months of antibiotic therapy, a history of persistent effusion for 3 or more months per episode of otitis media, more than 7 episodes of otitis media in 6 months, a bilateral conductive hearing loss of 20 dB or more and a persistently abnormal tympanic membrane. Some respondents were more likely to recommend tube insertion if there were parental concerns about hearing problems or the frequency or severity of episodes of otitis media. Otolaryngologists agreed about the role of tympanostomy tubes in 1 of 4 case vignettes but disagreed about whether adenoidectomy should also be performed in that instance. Most viewed tympanostomy tube insertion as beneficial, with few adverse effects. INTERPRETATION: There is a lack of consensus among practising otolaryngologists in Ontario as to which children with recurrent otitis media or persistent effusion should undergo bilateral myringotomy with tympanostomy tube insertion. These findings suggest the need to revisit clinical guidelines for this procedure.

Differences between men and women in the rate of use of hip and knee arthroplasty.

BACKGROUND: Previous studies suggest that, for some conditions, women receive fewer health care interventions than men. We estimated the potential need for arthroplasty and the willingness to undergo the procedure in both men and women and examined whether there were differences between the sexes. METHODS: All 48,218 persons 55 years of age or older in two areas of Ontario, Canada, were surveyed by mail and telephone to identify those with hip or knee problems. In these subjects, we assessed the severity of arthritis and the presence of coexisting conditions by questionnaire, documented arthritis by examination and radiography, and conducted interviews to evaluate the subjects' willingness to undergo arthroplasty. The potential need for arthroplasty was defined by the presence of severe symptoms and disability, the absence of any absolute contraindications to surgery, and clinical and radiographic evidence of arthritis. The estimates of need were then adjusted for the subjects' willingness to undergo arthroplasty. RESULTS: The overall response rates were at least 72 percent for the questionnaires and interviews. As compared with men, women had a higher prevalence of arthritis of the hip or knee (age-adjusted odds ratio, 1.76; P<0.001) and had worse symptoms and greater disability, but women were less likely to have undergone arthroplasty (adjusted odds ratio, 0.78; P<0.001). Despite their equal willingness to have the surgery, fewer women than men had discussed the possibility of arthroplasty with a physician (adjusted odds ratio, 0.63). The numbers of people with a potential need for hip or knee arthroplasty were 44.9 per 1000 among women and 20.8 per 1000 among men. After adjustment for willingness to undergo the procedure, the numbers were 5.3 per 1000 for women and 1.6 per 1000 for men. CONCLUSIONS: There is underuse of arthroplasty for severe arthritis in both sexes, but the degree of underuse is more than three times as great in women as in men.