Increasing the efficiency of Monte Carlo cohort simulations with variance reduction techniques.

The authors discuss techniques for Monte Carlo (MC) cohort simulations that reduce the number of simulation replications required to achieve a given degree of precision for various output measures. Known as variance reduction techniques, they are often used in industrial engineering and operations research models, but they are seldom used in medical models. However, most MC cohort simulations are well suited to the implementation of these techniques. The authors discuss the cost of implementation versus the benefit of reduced replications.     

Serum monitoring of antipsychotic drug levels during concomitant administration of sertraline and antipsychotic medication.

OBJECTIVE: To assess whether pharmacokinetic drug interactions occur when sertraline is added to antipsychotic medications. METHOD: Forty-eight patients with remitted DSM-IV schizophrenia and comorbid major depression were randomized to placebo for 6 weeks or sertraline 50 mg for 4 weeks followed by sertraline 50 mg to 100 mg for 2 weeks for nonresponders. Treatment with the patients' usual antipsychotic continued. Weekly clinical outcome assessments occurred for 6 weeks, and serum samples for drug monitoring were collected at Weeks 1, 5, and 6. Serum concentrations of sertraline and antipsychotics were measured with standard assays. RESULTS: In both placebo- and sertraline-treated groups, most patients displayed minor fluctuations in antipsychotic serum levels over 6 weeks. There was no clinical evidence of drug interactions in the sertraline-treated group. CONCLUSIONS: Clinically significant adverse effects did not occur despite variable antipsychotic serum levels with or without sertraline. Concern about pharmacokinetic interactions should not deter the use of sertraline for depression in individuals with schizophrenia.     

Academic Degrees and Clinical Practice Characteristics: The University of Washington Physician Assistant Program: 1969-2000

CONTEXT: The physician assistant profession has been moving toward requiring master's degrees for new practitioners, but some argue this could change the face of the discipline. PURPOSE: To see if there is an association between physician assistants' academic degrees and practice in primary care, in rural areas, and with the medically underserved. METHODS: Surveys were sent to 880 graduates of the first 32 University of Washington physician assistant classes through 2000. Respondents noted their academic degree at program entry and the highest degree attained at any time up to the time of survey. Relationships between practice characteristics and academic degree levels were tested by unadjusted odds ratios and logistic regression after controlling for year of graduation and sex. RESULTS: Of the 478 respondents, 54% worked in primary care, about 30% practiced in nonmetropolitan communities, and 42% reported providing care for the medically underserved. Respondents with no degree (33% of total at entry, 24% at survey) were significantly more likely than degree holders to work in primary care and nonmetropolitan areas. Respondents with no degree at program entry were significantly more likely, and those with no degree at the time of the survey were marginally more likely, to self-report work with the medically underserved. CONCLUSION: Respondents with no academic degree are significantly more likely to demonstrate a commitment to primary, rural, and underserved health care. These findings may inform the national debate about the impact of required advanced degrees on the practice patterns of nonphysician providers.     

Cause-specific and total mortality in the Canterbury (New Zealand) insulin-treated Diabetic Registry population: a 15-year follow-up study.

AIMS: To establish all-cause and cause-specific death rates, and risk factors for mortality in insulin-treated diabetic individuals living in the province of Canterbury, New Zealand. METHODS: Insulin-treated diabetic subjects (n = 995) on the Canterbury Diabetes Registry were followed up over 15 years and vital status determined. Death rates were standardized and hazard regression was used to model the effects of demographic covariates on relative survival time. RESULTS: There were 419 deaths in 11 226.3 person-years of follow-up with a standardized mortality ratio (SMR) of 2.0 (95% confidence interval (CI) 1.8-2.2). Relative mortality was greatest for the group aged 0-29 years (SMR 3.0 (95% CI 2.4-3.7)). After controlling for diabetes duration and gender, a 10-year increment in age of onset was associated with a 33% decrease in relative hazard (95% CI 29-36%), indicating that excess mortality due to diabetes declines with rising age of onset. After controlling for age of onset and gender, each 10-year increment in duration of diabetes is associated with a 26% decrease in relative hazard (95% CI 24-29%), indicating that with longer survival the mortality hazard approaches the general population hazard. Relative mortalities were increased for cardiovascular, renal and respiratory disease, but not malignancy. Relative mortality from acute metabolic complications was increased in the subgroup with age of onset of diabetes < 30 years and requiring insulin within 1 year of diagnosis. CONCLUSIONS: Mortality rates are high for insulin-treated diabetic individuals relative to the general population.     

Hospital Episode Statistics and trends in ophthalmic surgery 1998 – 2004

Background  

Hospital episode statistics (HES) is a UK national database for the National Health Service (NHS), now available online. The purpose of this study was to observe trends in ophthalmic operations performed during the period from 1998 to 2004, using this data.     

The influence of different error estimates in the detection of post-operative cognitive dysfunction using reliable change indices with correction for practice effects.

The reliable change index (RCI) expresses change relative to its associated error, and is useful in the identification of post-operative cognitive dysfunction (POCD). This paper examines four common RCIs that each account for error in different ways. Three rules incorporate a constant correction for practice effects and are contrasted with the standard RCI that had no correction for practice. These rules are applied to 160 patients undergoing coronary artery bypass graft (CABG) surgery who completed neuropsychological assessments preoperatively and 1 week post-operatively using error and reliability data from a comparable healthy non-surgical control group. The rules all identify POCD in a similar proportion of patients, but the use of the within subject standard deviation, expressing the effects of random error, as an error estimate is a theoretically appropriate denominator when a constant error correction, removing the effects of systematic error, is deducted from the numerator in a RCI.     

Reduction of embryonic intracellular pH: a potential mechanism of acetazolamide-induced limb malformations

The effects of acetazolamide on the developing rodent limb bud were postulated to result from a reduction of intracellular pH (pHi). Embryonic intracellular pH was calculated from transplacental distribution of the weak acid, 5,5'-dimethyloxazolidine-2,4-dione, in teratogenically sensitive (C57BL/6) and resistant (SWV) inbred mice. pHi was reduced by acetazolamide treatment in C57 embryos and limb buds but not in SWV samples. Acetazolamide teratogenesis can be exacerbated by coadministration of amiloride, presumably through inhibition of Na+/H+ exchange attributable to the latter agent. pHi reduction after such treatment was more profound than after acetazolamide alone, providing further support for the central hypothesis. pH was also reduced in other embryonic (embryo plasma) and extraembryonic compartments (exocoelomic fluid, amniotic fluid). pH changes in these compartments could also lead or contribute to abnormal development.     

Blood lead levels in children in Perth,Western Australia

Abstract: This study reports on an analysis of the lead concentrations in 123 venous blood samples collected from Perth children aged between two months and 17 years attending Princess Margaret Hospital. The overall geometric mean was 6.9 μg lead per 100 ml whole blood, with 95 per cent of results lying between 3.2 and 14.8 μg/100 ml. Among children under five years of age, those aged between 18 months and two years had the highest geometric mean blood lead (11.1 μg/100 ml). There were no consistent associations between geometric mean blood lead and area of residence, age group or sex. In this sample of Perth children, the mean blood lead concentration was lower than those reported in other studies. Less than 0.1 per cent of children of the age range studied would have been expected to have lead levels exceeding the NHMRC ‘level of concern’ (25 μg/100 ml) current at the time of the study. However, the recent adoption of goal of less than 10 μg/100 ml could mean that lead levels in up to 21 per cent of Perth children would now be regarded as excessive.