Association between waiting time for radiotherapy after surgery for early-stage breast cancer and survival outcomes in Ontario: a population-based outcomes study.

Background: After surgery for early-stage breast cancer (bca), adjuvant radiotherapy (rt) decreases the risk of locoregional recurrence and death from bca. It is unclear whether delays to the initiation of adjuvant rt are associated with inferior survival outcomes. Methods: This population-based retrospective cohort study included a random sample of 25% of all women with stage i or ii bca treated with adjuvant rt in Ontario between 1 September 2001 and 31 August 2002, when, because of capacity issues, wait times for radiation were abnormally long. Pathology reports were manually abstracted and deterministically linked to population-level administrative databases to obtain information about recurrence and survival outcomes. Cox proportional hazards modelling was used to evaluate the association between waiting time and survival outcomes. A composite survival outcome was used to ensure that all possible measurable harms of delay would be captured. The composite outcome, event-free survival, included locoregional recurrence, development of metastatic disease, and bca-specific mortality. Results: We identified 1028 women with stage i or ii bca who were treated with breast-conserving surgery and adjuvant rt. For the 599 women who were treated with adjuvant radiation without intervening chemotherapy, a waiting time of 12 weeks or more from surgery to the start of radiation appeared to be associated with worse event-free survival after a median follow-up of 7.2 years (hazard ratio for the composite outcome: 1.44; 95% confidence interval: 0.98 to 2.11; p = 0.07). For the 429 women who received intervening adjuvant chemotherapy, a waiting time of 6 weeks or more from completion of chemotherapy to start of radiation was associated with worse event-free survival after a median follow-up of 7.4 years (hazard ratio: 1.50; 95% confidence interval: 1.00 to 2.22; p = 0.047). Conclusions: Delay to the initiation of adjuvant rt after breast-conserving surgery is associated with inferior bca survival outcomes. The good prognosis for patients with early-stage bca limits the statistical power to detect an effect of delay to rt. Given that there is no plausible advantage to delay, we agree with Mackillop that time to initiation of rt should be kept "as short as reasonably achievable."

The predictors and economic burden of early-, mid- and late-onset cardiac implantable electronic device infections: a retrospective cohort study in Ontario, Canada.

The rate of cardiac implantable electronic device (CIED) infection is increasing with time. We sought to determine the predictors, relative mortality, and cost burden of early-, mid- and late-onset CIED infections. We conducted a retrospective cohort study of all CIED implantations in Ontario, Canada between April 2013 and March 2016. The procedures and infections were identified in validated, population-wide health-care databases. Infection onset was categorized as early (0-30 days), mid (31-182 days) and late (183-365 days). Cox proportional hazards regression was used to assess the mortality impact of CIED infections, with infection modelled as a time-varying covariate. A generalized linear model with a log-link and γ distribution was used to compare health-care system costs by infection status. Among 17 584 patients undergoing CIED implantation, 215 (1.2%) developed an infection, including 88 early, 85 mid, and 42 late infections. The adjusted hazard ratio (aHR) of death was higher for patients with early (aHR 2.9, 95% CI 1.7-4.9), mid (aHR 3.3, 95% CI 1.9-5.7) and late (aHR 19.9, 95% CI 9.9-40.2) infections. Total mean 1-year health costs were highest for late-onset (mean Can$113 778), followed by mid-onset (mean Can$85 302), and then early-onset (Can$75 415) infections; costs for uninfected patients were Can$25 631. After accounting for patient and procedure characteristics, there was a significant increase in costs associated with early- (rate ratio (RR) 3.1, 95% CI 2.3-4.1), mid- (RR 2.8, 95% CI 2.4-3.3) and late- (RR 4.7, 95% CI 3.6-6.2) onset infections. In summary, CIED infections carry a tremendous clinical and economic burden, and this burden is disproportionately high for late-onset infections.

Complications following surgery with or without radiotherapy or radiotherapy alone for prostate cancer.

BACKGROUND: Men undergoing treatment of clinically localised prostate cancer may experience a number of treatment-related complications, which affect their quality of life. METHODS: On the basis of population-based retrospective cohort of men undergoing surgery, with or without subsequent radiotherapy, or radiotherapy alone for prostate cancer in Ontario, Canada, we measured the incidence of treatment-related complications using administrative and billing data. RESULTS: Of 36 984 patients, 15 870 (42.9%) underwent surgery alone, 4519 (12.2%) underwent surgery followed by radiotherapy, and 16 595 (44.9%) underwent radiotherapy alone. For all end points except urologic procedures, the 5-year cumulative incidence rates were lowest in the surgery only group and highest in the radiotherapy only group. Intermediary rates were seen in the surgery followed by radiotherapy group, except for urologic procedures where rates were the highest in this group. Although age and comorbidity were important predictors, radiotherapy as the primary treatment modality was associated with higher rates for all complications (adjusted hazard ratios 1.6-4.7, P=0.002 to <0.0001). CONCLUSIONS: In patients treated for prostate cancer, radiation after surgery increases the rate of complications compared with surgery alone, though these rates remain lower than patients treated with radiation alone. This information may inform patient and physician decision making in the treatment of prostate cancer.

Health system costs for stage-specific breast cancer: a population-based approach.

OBJECTIVE: The objective of the present analysis was to determine the publicly funded health care costs associated with the care of breast cancer (bca) patients by disease stage. METHODS: Incident cases of female invasive bca (2005-2009) were extracted from the Ontario Cancer Registry and linked to administrative datasets from the publicly funded system. The type and use of health care services were stratified by disease stage over the first 2 years after diagnosis. Mean costs and costs by type of clinical resource used in the care of bca patients were compared with costs for a matched control group. The attributable cost for the 2-year time horizon was determined in 2008 Canadian dollars. RESULTS: This cohort study involved 39,655 patients with bca and 190,520 control subjects. The average age in those groups was 61.1 and 60.9 years respectively. Most bca patients were classified as either stage i (34.4%) or stage ii (31.8%). Of the bca cohort, 8% died within the first 2 years after diagnosis. The overall mean cost per bca case from a public payer perspective in the first 2 years after diagnosis was $41,686. Over the 2-year time horizon, the mean cost increased by stage: i, $29,938; ii, $46,893; iii, $65,369; and iv, $66,627. The attributable cost of bca was $31,732. Cost drivers were cancer clinic visits, physician billings, and hospitalizations. CONCLUSIONS: Costs of care increased by stage of bca. Cost drivers were cancer clinic visits, physician billings, and hospitalizations. These data will assist planning and decision-making for the use of limited health care resources.

Population-based utilization of radiation therapy by a Canadian breast cancer cohort.

We examined trends in radiation therapy (rt) utilization by a population-based breast cancer cohort in Ontario. The provincial cancer registry provided a breast cancer cohort based on diagnosis dates from April 1, 2005, to March 31, 2010. Staging information was also available. The cohort was then linked, by encrypted health card number, to linkable administrative datasets, including rt utilization. The average age in the identified female breast cancer cohort (n = 39,656) was 61.6 ± 14.0 years. Almost two thirds of the patients (n = 25,225) received rt, and staging information was available for 22,988 patients (9541 stage i, 8516 stage ii, 4050 stage iii, and 881 stage iv). The average number of rt courses received by the patients was 1.4 ± 0.7 for stage i, 1.8 ± 1.1 for stage ii, 2.5 ± 1.3 for stage iii, and 2.8 ± 2.4 for stage iv. The ratio of conventional rt to intensity-modulated rt was 70.9%:16.6% for stage i, 71.6%:11.3% for stage ii, 74.6%:4.6% for stage iii, and 89.6%:2.2% for stage iv. From 2005 to 2010, almost two thirds of a Canadian female breast cancer cohort received rt, and the average number of courses increased with disease severity. A similar trend was observed with the type of rt (use of conventional rt increased with disease severity). The next step is to apply unit costs to the number of fractions and to obtain rt planning and radiation therapist times.

Age at diagnosis predicts local recurrence in women treated with breast-conserving surgery and postoperative radiation therapy for ductal carcinoma in situ: a population-based outcomes analysis.

PURPOSE: The main goal of treating ductal carcinoma in situ (dcis) is to prevent the development of invasive breast cancer. Most women are treated with breast-conserving surgery (bcs) and radiotherapy. Age at diagnosis may be a risk factor for recurrence, leading to concerns that additional treatment may be necessary for younger women. We report a population-based study of women with dcis treated with bcs and radiotherapy and an evaluation of the effect of age on local recurrence (lr). METHODS: All women diagnosed with dcis in Ontario from 1994 to 2003 were identified. Treatments and outcomes were collected through administrative databases and validated by chart review. Women treated with bcs and radiotherapy were included. Survival analyses were performed to evaluate the effect of age on outcomes. RESULTS: We identified 5752 cases of dcis; 1607 women received bcs and radiotherapy. The median follow-up was 10.0 years. The 10-year cumulative lr rate was 27% for women younger than 45 years, 14% for women 45-50 years, and 11% for women more than 50 years of age (p < 0.0001). The 10-year cumulative invasive lr rate was 22% for women younger than 45 years, 10% for women 45-50 years, and 7% for women more than 50 years of age (p < 0.0001). On multivariate analyses, young age (<45 years) was significantly associated with lr and invasive lr [hazard ratio (hr) for lr: 2.6; 95% confidence interval (ci): 1.9 to 3.7; p < 0.0001; hr for invasive lr: 3.0; 95% ci: 2.0 to 4.4; p < 0.0001]. An age of 45-50 years was also significantly associated with invasive lr (hr: 1.6; 95% ci: 1.0 to 2.4; p = 0.04). CONCLUSIONS: Age at diagnosis is a strong predictor of lr in women with dcis after treatment with bcs and radiotherapy.

Sociodemographic factors associated with cervical cancer screening coverage and follow-up of high grade abnormal results in a population-based cohort.

INTRODUCTION: A cervical cytology based screening program is effective if there is regular screening of the 'at risk' population and close follow-up of those labeled abnormal. METHODS: This is a population cohort study of women between 20-69 year old who were eligible in Ontario from 2008-2010. We used administrative data to evaluate the rates of cervical cancer screening and follow-up of high grade Pap tests. Variation in cervical cytology coverage and follow-up of high grade abnormal results are associated with age, area level income and health region. Multivariate logistic regression was used to identify independent factors associated with screening and followup. RESULTS: 3.7million women were eligible for screening of which 72% had a Pap smear in the prior 3years. These rates varied by age, income and region (p<0.0001). Women residing in the lowest income neighborhoods were half as likely to be screened (p<0.0001). 83% of those with an high grade intraepithelial lesion Pap test result had follow-up with colposcopy or treatment within 6months and this varied by year, age, income and region (p<0.0001). CONCLUSIONS: Despite universal health coverage, cervical cancer screening rates are suboptimal with older and low income women being at greatest risk. Follow-up among women with high grade abnormal tests is mediocre at 3months and acceptable at 6months. Novel models of cervical cancer screening program implementation are needed to address these inadequacies.

HER2/neu and Ki-67 expression predict non-invasive recurrence following breast-conserving therapy for ductal carcinoma in situ.

BACKGROUND: Ductal carcinoma in situ (DCIS) is a non-invasive form of breast cancer that may progress to invasive cancer. Identification of factors that predict recurrence and distinguish DCIS from invasive recurrence would facilitate treatment recommendations. We examined the prognostic value of nine molecular markers on the risks of local recurrence (DCIS and invasive) among women treated with breast-conserving therapy. METHODS: A total of 213 women who were treated with breast-conserving therapy between 1982 and 2000 were included; 141 received breast-conserving surgery alone and 72 cases received radiotherapy. We performed immunohistochemical staining on the DCIS specimen for nine markers: oestrogen receptor, progesterone receptor, Ki-67, p53, p21, cyclinD1, HER2/neu, calgranulin and psoriasin. We performed univariable and multivariable survival analyses to identify markers associated with the recurrence. RESULTS: The rate of recurrence at 10 years was 36% for patients treated with breast-conserving surgery alone and 18% for women who received breast-conserving surgery and radiotherapy. HER2/neu+/Ki-67+ expression was associated with an increased risk of DCIS recurrence, independent of grade and age (HR=3.22; 95% CI: 1.47-7.03; P=0.003). None of the nine markers were predictive of invasive recurrence. CONCLUSION: Women with a HER2/neu/neu+/Ki67+ DCIS have a higher risk of developing DCIS local recurrence after breast-conserving surgery.

Expression of HER2neu in ductal carcinoma in situ is associated with local recurrence.

AIMS: Determination of the risk of recurrence after local excision of ductal carcinoma in situ (DCIS) remains a challenge. Molecular profiling based on immunohistochemical staining to oestrogen receptor (ER), progesterone receptor (PR) and HER2neu improved risk prediction in invasive breast cancer, but few studies have evaluated if molecular classification of DCIS predicts local recurrence. We evaluated the expression of ER, PR and HER2neu in DCIS to determine if molecular classification predicts local recurrence after breast-conserving therapy for DCIS. MATERIALS AND METHODS: We reviewed the records of patients with DCIS treated between 1987 and 2000, carried out a pathology review and immunohistochemical staining for ER, PR and HER2neu and categorised cases into four molecular phenotypes [luminal A (ER+ and/or PR+, HER2neu-), luminal B (ER+ and/or PR+, HER2neu+), HER2neu subtype (ER-, PR-, HER2neu+), triple negative (ER-, PR-, HER2neu-)]. We evaluated the association between the molecular subtype and the development of local recurrence. RESULTS: In total, 180 cases of DCIS were included in the study (luminal A, n=113; luminal B, n=25; HER2neu type, n=29; triple negative, n=13). The median follow-up time was 8.7 years. We observed higher rates of local recurrence among luminal B (40%) and HER2neu type (38%) DCIS compared with luminal A (21%) and triple negative (15%) DCIS. On multivariable analysis, HER2neu overexpression was associated with an increased risk of local recurrence (hazard ratio=1.98; 95% confidence interval: 1.11, 3.53, P=0.02). CONCLUSION: HER2neu expression in DCIS is a significant predictor of local recurrence, whereas luminal A and triple negative phenotypes are associated with relatively low risks of local recurrence.

Population-based home care services in breast cancer: utilization and costs.

OBJECTIVE: To determine utilization and costs of home care services (hcs) for individuals with a diagnosis of breast cancer (bc). METHODS: Incident cases of invasive bc in women were extracted from the Ontario Cancer Registry (2005-2009) and linked with other Ontario health care administrative databases. Control patients were selected from the population of women never diagnosed with any type of cancer. The types and proportions of hcs used were determined and stratified by disease stage. Attributable home care utilization and costs for bc patients were determined. Factors associated with hcs costs were assessed using regression analysis. RESULTS: Among the 39,656 bc and 198,280 control patients identified (median age: 61.6 years for both), 75.4% of bc patients used hcs (62.1% stage i; 85.7% stage ii; 94.6% stage iii; 79.1% stage iv) compared with 14.6% of control patients. The number of hcs used per patient-year were significantly higher for the bc patients than for the control patients (14.97 vs. 6.13, p < 0.01), resulting in higher costs per patient-year ($1,210 vs. $325; $885 attributable cost to bc, p < 0.01). The number of hcs utilized and the associated costs increased as the bc stage increased. In contrast, hcs costs decreased as income increased and as previous health care exposure decreased. INTERPRETATION: Patients with bc used twice as many hcs, resulting in costs that were almost 4 times those observed in a matched control group. Less than an additional $1000 per bc patient per year were spent on hcs utilization in the study population.

End of life care for women with gynecologic cancers.

BACKGROUND: Little is known about patterns of end of life (EOL) care in gynecologic cancer patients. This paper reports on five EOL quality indicators: (1) chemotherapy in last 2 weeks of life (2) death in an acute care bed (3) emergency department visits in last 2 weeks of life (4) home care (nursing) visits in last 6 months of life (5) physician house calls in last 2 weeks of life. METHODS: A population-based, retrospective cohort study using administrative sources of health care data which was conducted as part of the Project for an Ontario Women's Health Report Card. It describes five health services received near the EOL by women who died of ovarian, uterine or cervical cancer in 2003-2004 in Ontario, Canada. Measures were stratified by age, income and region. RESULTS: The cohort included 2040 women. Four percent received chemotherapy, 34% visited the emergency department; 27% received a physician house call; 73% received a home care visit; and 51% died in an acute care bed. Older age was associated with lower use of each service. Living in a lower income neighborhood was associated with lower physician home visits. Regional variation across the province was observed for 3 indicators. INTERPRETATION: Observations made in this study can be used to inform interventions to improve EOL care for women with gynecological cancers. Tracking indicators over time serves to monitor response to improvement interventions. Reporting on the specific needs of this population helps assure that gaps in this domain of care are addressed.

Patterns of care for the initial management of cervical cancer in Ontario.

OBJECTIVE: To facilitate the planning of resources for cancer services in Ontario, Cancer Care Ontario commissioned an evaluation of operative services delivered for cervical cancer. METHODS: Women with an incident diagnosis of cervical cancer were identified from 1 April, 2003 to 31 March, 2004 using the Ontario Cancer Registry. Record linkages were created to other provincial health databases such as the Ontario Health Insurance Plan. RESULTS: There were 513 incident cases. Disease-specific rates of cancer were higher in rural areas and those from lower income quintiles. Forty-three percent of women had no surgery. Use of surgery did not appear to vary by SEC, urban/rural residence or LHIN. Women of younger age were more like to receive surgery for cervical cancer. Gynecologists conducted 63% of the operations. Gynecologics were most likely to complete a lymphadenectomy (70.3%). All women were assessed by CXR. Only 22% of women had a CT scan of the abdomen and pelvis. Radiation consults were performed in half of the women with cervix cancer but treatment was only delivered to half of those seen. Medical oncologists saw about 10% of women with cervical cancers. CONCLUSIONS: There appear to be variations in incidence rates of cervical cancer, with cancers being more frequent in rural areas. In two-thirds of the population, surgery is performed in the region where the patient lives. Subspecialty care from gynecologic oncologists was provided to one-third of women. These preliminary data would be enhanced with further information such as comorbidity, treatment intent (palliative/curative), histology, grade and stage.

Patterns of care in the initial management of women with vulvar cancers in Ontario.

BACKGROUND: To facilitate the planning of future resources for cancer services in Ontario, Cancer Care Ontario commissioned an evaluation of operative services delivered for vulvar cancer. METHODS: Women with an incident diagnosis of vulvar malignancy were identified from 1, April 2003 to 31 March, 2004 using the Ontario Cancer Registry. Record linkages were created to other provincial health databases such as the Ontario Health Insurance Plan. RESULTS: Vulvar cancers affected 148 women. Disease specific rates of cancer were higher in rural areas and in women in the lower income quintiles. No surgery occurred in 17.6% of women. Use of surgery did not appear to vary by urban/rural residence or LHIN. Ontario's 17 gynecologic oncologists performed 75% of the surgeries. Groin lymphadenectomy rate was 52.8%. Surgery was performed in the LHIN of residence for 41% of women. All women were assessed by CXR. CT scan of the abdomen and pelvis occurred in 77%. MRIs were done infrequently. Radiation consults were preformed in half of the women with vulvar cancer but treatment was only delivered in half of those seen. Medical oncologists saw about 10% of women with gynecologic cancers. CONCLUSIONS: There appear to be variations in incidence rates of vulvar cancer with disease being more frequent in rural areas. Subspecialty care from gynecologic oncologists was provided to 75% of women. Rates of lymphadenectomy as part of a surgical attempt occurred in 52.8% of women. These data would be enhanced with further information such as comorbidity, treatment intent (palliative/curative), histology, grade and stage.

Patterns of care in the initial management of women with ovarian cancer in Ontario.

BACKGROUND: To facilitate the planning of future resources for cancer services in Ontario, Cancer Care Ontario commissioned an evaluation of operative services delivered for ovarian cancers. The affected population was characterized in terms of age, location of residence, and SES. Operative care delivery was described in terms of inpatient verses outpatient access, LHIN of treatment, surgical specialist providing treatment, and specific operative procedures. The investigations and consults around the time of diagnosis are described. METHODS: Women with an incident diagnosis of an ovarian malignancy were identified from 1 April 2003 to 31 March 2004 using the Ontario Cancer Registry. Record linkages were created to other provincial health databases such as the Ontario Health Insurance Plan. RESULTS: We report on 963 women with ovarian cancer. The incidence of disease was related to increasing age. Access to surgery correlated with the highest income quintile, urban residence and LHIN. Twenty-seven percent of women did not have surgery for their ovarian cancer. Women of younger age were more like to receive surgery for ovarian cancer. Use of a laparotomy for biopsy was most common in community hospital (40%). Lymphadenectomy rates were low overall; rates for gynecologic oncologists were 13.2%. All women were assessed by CXR. CT scan of abdomen and pelvis occurred in 77% of women. MRIs were done infrequently. Medical oncology were involved in 26.6% of the patients. CONCLUSIONS: These pilot data would be enhanced with further information such as comorbidity, treatment intent (palliative/curative), histology, grade and stage. However, there are clear referral patterns to academic centres which means a need for manpower and hospital resources to deal with this population.

Patterns of surgical care for uterine cancers in Ontario.

OBJECTIVES: To facilitate the planning of future resources for cancer services in Ontario, Cancer Care Ontario commissioned an evaluation of operative services delivered for uterine cancer. METHODS: Women with an incident diagnosis of a uterine malignancy were identified from 1 April 2003 to 31 March 2004 using the Ontario Cancer Registry. Record linkages were created to other provincial health databases such as the Ontario Health Insurance Plan. RESULTS: Uterine cancer affected 1,436 women. Disease specific rates of cancer were higher in rural areas and those from the highest income quintiles. Surgery occurred in 94.7% of women. Use of surgery did not appear to vary by SEC, urban/rural residence or LHIN. Gynecologists conducted 76.1% of the operations. Lymphadenectomy took place in 18.7% of women. Lymphadenectomy rates were highest in gynecologic oncologists (43.3%). All women were assessed by CXR. Radiation therapy consults were preformed in half of the women with uterine cancer but treatment was only delivered in half of those seen. Medical oncologists saw about 6.3% of women with uterine cancers. CONCLUSIONS: There appear to be variations in incidence rates of uterine cancer with disease being more frequent in those of the highest SES. In two-thirds of the population, surgery is delivered in the region where the patient lives. Subspecialty care from gynecologic oncologists was provided to one-third of women. Rates of lymphadenectomy as part of a surgical attempt to assess disease spread appear low. These pilot data would be enhanced with further information such as comorbidity, treatment intent (palliative/curative), histology, grade and stage.

Risk factors for grade 3-4 liver enzyme elevation in HIV and hepatitis C coinfected patients on combination antiretroviral therapy.

Although coinfection with hepatitis C (HCV) is an established risk factor for hepatotoxicity in HIV-positive patients receiving combination antiretroviral therapy (cART), specific variables that may be predictive of severe hepatotoxicity among co-infected patients receiving cART remain poorly defined. A retrospective cohort study of HIV/HCV coinfected adults from two HIV treatment centers covering the period between December 1998 and December 2003 was conducted to address this question. The primary endpoint of the study was the occurrence of grade 3 or 4 elevation of serum alanine aminotransferase (ALT) during follow-up and the primary predictors of interest were specific antiretrovirals. One hundred five coinfected patients receiving cART for a median of 70 months (interquartile range [IQR], 37, 83) were included in the analysis. Twenty-three (22%) patients developed a grade 3 or 4 increase in serum ALT at least once in follow-up. In univariate analysis, current receipt of lopinavir/ritonavir (LPV/r) (odds ratio [OR] 3.09, 95% confidence interval [CI] 1.14-8.34, p = 0.03), baseline ALT (OR 1.01, 95% CI 1.00-1.02, p = 0.004), and current use of boosting ritonavir (OR 2.84, 95% CI 1.16-7.00, p = 0.02) were significantly associated with a grade 3 or 4 increase in serum ALT, although most patients receiving boosting ritonavir were on lopinavir/ritonavir based regimens. Patients receiving LPV/r had been previously exposed to significantly more antiretrovirals (p < 0.0001), protease inhibitors (p < 0.0001), and nucleoside analogues (p = 0.0009) compared to the rest of the cohort. Further research to better clarify risk factors for hepatotoxicity in coinfected patients is warranted given the challenges in treating this population.

Novel murine model of pneumococcal pneumonia: use of temperature as a measure of disease severity to compare the efficacies of moxifloxacin and levofloxacin.

Surface temperature measured by an infrared temperature-scanning thermometer was used to evaluate disease severity and predict imminent death in a murine model of pneumococcal pneumonia. We showed that a decrease in temperature was associated with increasing severity of disease and concomitant histological changes and also that a temperature of 30 degrees C or less was a predictor of death. Furthermore, viable bacterial counts in the lungs of mice euthanized at a temperature of < or = 30 degrees C were not significantly different from those seen in the lungs of mice allowed to die without intervention. These data support temperature change as a more subtle indicator of outcome than death and demonstrate that this could be used as a reliable end point for euthanasia. To test the utility of our model in a drug trial, we examined the efficacies of moxifloxacin and levofloxacin by using temperature as a measure of disease severity prior to and during treatment. Regardless of the antibiotic used, mice assessed as moderately ill (temperature > or = 32 degrees C) at the start of treatment had better clinical and bacteriological outcomes than mice assessed as severely ill (temperature < 32 degrees C). However, moxifloxacin offered better protection and greater bacterial clearance than did levofloxacin in all infected mice independent of disease severity. This model not only allows a more subtle evaluation of drug efficacy but also ensures a better degree of standardization and a more humane approach to drug efficacy studies involving animals.

Case-control analysis with partial knowledge of exposure misclassification probabilities.

Consider case control analysis with a dichotomous exposure variable that is subject to misclassification. If the classification probabilities are known, then methods are available to adjust odds-ratio estimates in light of the misclassification. We study the realistic scenario where reasonable guesses, but not exact values, are available for the classification probabilities. If the analysis proceeds by simply treating the guesses as exact, then even small discrepancies between the guesses and the actual probabilities can seriously degrade odds-ratio estimates. We show that this problem is mitigated by a Bayes analysis that incorporates uncertainty about the classification probabilities as prior information.