RESUMO
This research investigates the performance of a proportion-based approach to meta-analytic moderator estimation through a series of Monte Carlo simulations. This approach is most useful when the moderating potential of a categorical variable has not been recognized in primary research and thus heterogeneous groups have been pooled together as a single sample. Alternative scenarios representing different distributions of group proportions are examined along with varying numbers of studies, subjects per study, and correlation combinations. Our results suggest that the approach is largely unbiased in its estimation of the magnitude of between-group differences and performs well with regard to statistical power and type I error. In particular, the average percentage bias of the estimated correlation for the reference group is positive and largely negligible, in the 0.5-1.8% range; the average percentage bias of the difference between correlations is also minimal, in the -0.1-1.2% range. Further analysis also suggests both biases decrease as the magnitude of the underlying difference increases, as the number of subjects in each simulated primary study increases, and as the number of simulated studies in each meta-analysis increases. The bias was most evident when the number of subjects and the number of studies were the smallest (80 and 36, respectively). A sensitivity analysis that examines its performance in scenarios down to 12 studies and 40 primary subjects is also included. This research is the first that thoroughly examines the adequacy of the proportion-based approach. Copyright © 2012 John Wiley & Sons, Ltd.
RESUMO
OBJECTIVE: To determine the predictive value of commonly used clinical and laboratory factors for mortality in patients with pneumonia in the Arab world. METHODS: We retrospectively analyze the data collected from all inpatients over the age of 16 years with a diagnosis of pneumonia in Tawam Hospital, Al-Ain, United Arab Emirates between the years 1997 and 2002. Patients were grouped into those who survived and those who died in the hospital. Clinical and laboratory factors on admission were used to predict outcome using simple, and multiple logistic regression analyses. RESULTS: Among the 236 patients admitted, 122 were females (age 56.9 +/- 23 years), and 114 males (age 58.5 +/- 23 years). The 30-day mortality rate was 10%. The most common comorbid risk factors were diabetes mellitus in 23.7% and chronic obstructive pulmonary disease in 19.5%. Of the 236 patients, 145 had sputum culture on admission. Simple logistic regression analysis showed increasing age, presence of comorbidity, low systolic blood pressure, confused mental status, low serum albumin, high serum creatinine, raised blood urea nitrogen and raised partial pressure of carbon dioxide at the time of admission were associated with higher mortality. On the Stepwise-multilogistic regression analysis, the most significant factors influencing mortality were: older age, altered mental status, low systolic blood pressure, low serum albumin and raised serum creatinine. Using a scoring system developed in the presence or absence of these risk factors, a score of >=100 predicted high risk for mortality. CONCLUSION: The in-hospital mortality rate for pneumonia was 10%. Older age, altered mental status, low systolic blood pressure, low serum albumin concentration and raised serum creatinine at admission were predictive of poor outcome in this cohort of patients.
