Bruce A. Craig

A cost-effectiveness and cost-utility analysis of radiosurgery vs. resection for single-brain metastases

PURPOSE The median survival of well-selected patients with single-brain metastases treated with whole-brain irradiation and resection or radiosurgery is comparable, although a randomized trial of these two modalities has not been performed. In this era of cost containment, it is imperative that health-care professionals make fiscally prudent decisions. The present environment necessitates a critical appraisal of apparently equi-efficacious therapeutic modalities, and it is within this context that we present a comparison of the actual costs of resection and radiosurgery for brain metastases. METHODS AND MATERIALS Survival and quality of life outcome data for radiation alone or with surgery were obtained from two randomized trials, and radiosurgical results were obtained from a multiinstitutional analysis that specifically evaluated patients meeting surgical criteria. Only linear accelerator radiosurgery data were considered. Cost analysis was performed from a societal view point, and the following parameters were evaluated: actual cost, cost ratios, cost effectiveness, incremental cost effectiveness, cost utility, incremental cost utility, and national cost burden. The computerized billing records for all patients undergoing resection or radiosurgery for single-brain metastases from January 1989 to July 1994 were reviewed. A total of 46 resections and 135 radiosurgery procedures were performed. During the same time period, 454 patients underwent whole-brain radiation alone. An analysis of the entire bill was performed for each procedure, and each itemized cost was assigned a proportionate figure. The relative cost ratios of resection and radiosurgery were compared using the Wilcoxon rank sum test. Cost effectiveness of each modality, defined as the cost per year of median survival, was evaluated. Incremental cost effectiveness, defined as the additional cost per year of incremental gain in median survival, compared to the next least expensive modality, was also determined. To calculate the societal or national impact of these practices, the proportion of patients potentially eligible for aggressive management was estimated and the financial impact was determined using various utilization ratios for radiosurgery and surgery. RESULTS Both resection and radiosurgery yielded superior survival and functional independence, compared to whole brain radiotherapy alone, with minor differences in outcome between the two modalities; resection resulted in a 1.8-fold increase in cost, compared to radiosurgery. The latter modality yielded superior cost outcomes on all measures, even when a sensitivity analysis of up to 50% was performed. A reversal estimate indicated that in order for surgery to yield equal cost effectiveness, its cost would have to decrease by 48% or median survival would have to improve by 108%. The average cost per week of survival was

Cost effectiveness of highly active antiretroviral therapy in HIV-infected patients

310 for radiotherapy,

Comparison of self-reported, measured, metabolizable energy intake with total energy expenditure in overweight teens

524 for resection plus radiation, and

Incident use and outcomes associated with potentially inappropriate medication use in older adults

270 for radiosurgery plus radiation. CONCLUSIONS For selected patients, aggressive strategies such as resection or radiosurgery are warranted, as they result in improved median survival and functional independence. Radiosurgery appears to be the more cost-effective procedure.

Cyclooxygenase inhibitors in urinary bladder cancer: in vitro and in vivo effects

BACKGROUND Highly active antiretroviral therapy (HAART) has become the most important strategy for treating HIV infection in developed countries; however, access to HAART might vary under different funding policies. The Swiss health care system provides unrestricted access to HAART for all patients who need these newer combination therapies. This study investigated the impact of this funding policy on the society and health care system. METHODS A cost-effectiveness analysis with natural history data and productivity estimates was based on the Swiss HIV Cohort Study. A random sample of patient charts was used to estimate health care costs. In addition to a base-case scenario, a pessimistic and an optimistic scenario of natural disease history was developed. Costs were expressed in 1997 Swiss francs (100 CHF correspond to about US

Nonlinear Mixed Effects Model for Swine Growth

67) and effects as projected years of life gained. RESULTS In the analysis limited to health care costs, on the basis of projected survival in each scenario, the cost-effectiveness ratio was 33,000 CHF (base case), 14,000 CHF (optimistic), and 45,000 CHF (pessimistic) per year of life gained. When changes in productivity were included, cost savings occurred in the base-case and optimistic scenarios. The cost-effectiveness ratio was 11,000 CHF per year of life gained in the pessimistic scenario. CONCLUSIONS HAART increases expected survival and health care costs. However, when productivity gains are included, society will probably save costs or pay a low price for substantial health benefits. The study provides strong arguments, from a societal perspective, to continue the current policy of providing unrestricted access to HAART in Switzerland. The presented results also suggest that this policy could be of interest for other developed countries. Decision makers in developed countries where access to HAART is limited should re-evaluate their policy for the benefit of the society at large.

Evaluation of Alternative Nonlinear Mixed Effects Models of Swine Growth

BACKGROUND The accuracy of dietary energy assessment tools is critical to understanding the role of diet in the increasing rate of obesity. OBJECTIVES The purposes of our study in overweight adolescent boys and girls were 1) to assess the energy reporting bias of diet records against the referent of total energy expenditure (TEE) and 2) to compare the methods of determining energy needs by using measured metabolizable energy intake (MEI) and TEE. DESIGN Twenty girls [12-15 y, body mass index (in kg/m2) = 33.0 +/- 5] and 14 boys (12-14 y, body mass index = 27.4 +/- 4) participated in 2- to 3-wk metabolic balance studies. TEE was measured by using doubly labeled water (TEE(DLW)), and MEI was measured by bomb calorimetry of composite daily diet, urine, and fecal collections. Food records were collected before each study. RESULTS Food records underreported TEE(DLW) by 35 +/- 20%. Underreporting of energy intake was correlated with all macronutrient intake concentrations (g or kcal) (P < 0.0001). A multiple regression model showed that 86.4% of the variance in underreporting error was explained by dietary fat (g), BMI, and sex. The intrasubject CV was 3.9% for TEE(DLW) and 9.9% for MEI. MEI for weight stability (MEI(wtstb)) averaged 99 +/- 11% of TEE. CONCLUSIONS The increased underreporting of dietary intake with increasing body weight in teens may explain in part previous reports noting that there has been an increased incidence of obesity, although energy intakes have not appeared to increase. MEI(wtstb) and TEE(DLW) gave similar estimates of energy needs. This trial was registered at clinicaltrials.gov as NCT 00592137.

Estimating AIDS-free survival in a severely immunosuppressed asymptomatic HIV-infected population in the era of antiretroviral triple combination therapy. Swiss HIV Cohort Study.

BACKGROUND Most studies of potentially inappropriate medication (PIM) use among older adults have focused on prevalence rather than incidence. OBJECTIVES The goals of this study were to determine the 1-year incidence of PIM use among elderly Indiana Medicaid residents of nursing homes and to examine associations between incident PIM use and hospitalization and mortality. METHODS A retrospective analysis was conducted using Indiana Medicaid enrollment and administrative claims files. Individuals were included if they were Medicaid eligible and aged ≥65 years as of January 2003 and received nursing home services in each month of 2003 or until death in 2003. Individuals also had to receive nursing home services from October 2002 through December 2002 for inclusion in the sample. To focus analysis on incident PIM use, individuals who received any PIM prescription medication from October 2002 through December 2002 were excluded from the sample, as were those not prescribed any new medication in 2003. PIMs were identified using the 2003 Beers criteria. Associations between incident PIM use and hospitalization and mortality were assessed using logistic regression models after controlling for other risk factors. Potential selection bias was examined using bivariate probit models. RESULTS The study sample consisted of 7594 individuals (mean age, 83.07 years). A majority of the sample was female (76.5%), white (89.7%), and widowed (58.8%). Most individuals received care in nursing homes located in urban areas (5306 [69.9%]) and in the central region of Indiana (2838 [37.4%]). One-year incidence of PIM use was 42.1%. Incident PIM users were more likely to be hospitalized (odds ratio [OR] = 1.27; 95% CI, 1.10-C1.46) and more likely to die (OR = 1.46; 95% CI, 1.31-C1.62) in the 12 months after first receiving a PIM than nonusers, even after adjusting for demographic and clinical risk factors. CONCLUSIONS Incident PIM use was high among these elderly Indiana Medicaid residents of nursing homes. Individuals who began use of a PIM were at a higher risk of hospitalization and of dying.

Gene Expression Data: The Technology and Statistical Analysis

More than 14,000 people die from invasive transitional cell carcinoma (TCC) of the urinary bladder yearly in the United States. Cyclooxygenase (COX)-inhibiting drugs are emerging as potential antitumor agents in TCC. The optimal in vitro or in vivo systems to investigate COX inhibitor antitumor effects have not been defined. The purpose of this study was to determine COX-1 and COX-2 expression and antitumor effects of COX inhibitors in human TCC cell lines (HT1376, RT4, and UMUC3 cells) and xenografts derived from those cell lines. COX-2 expression (Western blot, immunocytochemistry) was high in HT1376, modest in RT4, and absent in UMUC3 cells in vitro. Similarly, COX-2 expression was noted in RT4 but not UMUC3 xenografts. COX-2 expression in HT1376 xenografts was slightly lower than that observed in vitro. None of four COX inhibitors evaluated (celecoxib, piroxicam, valeryl salicylate, and NS398) reduced TCC growth in standard in vitro proliferation assays at concentrations that could be safely achieved in vivo (≤5 μmol/L). Higher celecoxib concentrations (≥50 μmol/L) inhibited proliferation and induced apoptosis in all three cell lines. Celecoxib or piroxicam treatment in athymic mice significantly delayed progression of HT1376 xenografts, which express COX-2, but not UMUC3 xenografts that lack COX-2 expression. In conclusion, standard in vitro assays were not useful in predicting COX inhibitor antitumor effects observed in vivo. Athymic mice bearing TCC xenografts provide a useful in vivo system for COX inhibitor studies. Results of this study provide justification for further evaluation of COX inhibitors as antitumor agents against TCC. [Mol Cancer Ther 2006;5(2):329–36]

Associations Between BMI and Fat Taste Sensitivity in Humans

Abstract A mixed effects model version of a common swine growth function is introduced. This version, in which the mature BW of each pig is considered random, accounts for increasing variation with age and serial correlations across age that are common in serial BW growth data sets, thereby providing a much better fit and significantly smaller parameter SE. The inclusion of random effects also reduces the impact of selective sampling, which occurs when fast-growing pigs are removed for marketing, and provides a statistic (i.e., variance estimate) that describes the between-pig variation within a group. Comparisons between the fixed and mixed effects models are made using swine growth data consisting of 93 pigs with biweekly BW data from 54 to 138 d of age. In addition to a much better fit, the mixed effects model version is easily adaptable to stochastic modeling because only an additional random effect (i.e., mature BW) for each pig needs to be generated.