Rahim Moineddin

A simulation study of sample size for multilevel logistic regression models

BackgroundMany studies conducted in health and social sciences collect individual level data as outcome measures. Usually, such data have a hierarchical structure, with patients clustered within physicians, and physicians clustered within practices. Large survey data, including national surveys, have a hierarchical or clustered structure; respondents are naturally clustered in geographical units (e.g., health regions) and may be grouped into smaller units. Outcomes of interest in many fields not only reflect continuous measures, but also binary outcomes such as depression, presence or absence of a disease, and self-reported general health. In the framework of multilevel studies an important problem is calculating an adequate sample size that generates unbiased and accurate estimates.MethodsIn this paper simulation studies are used to assess the effect of varying sample size at both the individual and group level on the accuracy of the estimates of the parameters and variance components of multilevel logistic regression models. In addition, the influence of prevalence of the outcome and the intra-class correlation coefficient (ICC) is examined.ResultsThe results show that the estimates of the fixed effect parameters are unbiased for 100 groups with group size of 50 or higher. The estimates of the variance covariance components are slightly biased even with 100 groups and group size of 50. The biases for both fixed and random effects are severe for group size of 5. The standard errors for fixed effect parameters are unbiased while for variance covariance components are underestimated. Results suggest that low prevalent events require larger sample sizes with at least a minimum of 100 groups and 50 individuals per group.ConclusionWe recommend using a minimum group size of 50 with at least 50 groups to produce valid estimates for multi-level logistic regression models. Group size should be adjusted under conditions where the prevalence of events is low such that the expected number of events in each group should be greater than one.

Age- and sex-related prevalence of diabetes mellitus among immigrants to Ontario, Canada

Background: The majority of immigrants to Canada originate from the developing world, where the most rapid increase in prevalence of diabetes mellitus is occurring. We undertook a population-based study involving immigrants to Ontario, Canada, to evaluate the distribution of risk for diabetes in this population. Methods: We used linked administrative health and immigration records to calculate age-specific and age-adjusted prevalence rates among men and women aged 20 years or older in 2005. We compared rates among 1 122 771 immigrants to Ontario by country and region of birth to rates among long-term residents of the province. We used logistic regression to identify and quantify risk factors for diabetes in the immigrant population. Results: After controlling for age, immigration category, level of education, level of income and time since arrival, we found that, as compared with immigrants from western Europe and North America, risk for diabetes was elevated among immigrants from South Asia (odds ratio [OR] for men 4.01, 95% CI 3.82–4.21; OR for women 3.22, 95% CI 3.07–3.37), Latin America and the Caribbean (OR for men 2.18, 95% CI 2.08–2.30; OR for women 2.40, 95% CI: 2.29–2.52), and sub-Saharan Africa (OR for men 2.31, 95% CI 2.17–2.45; OR for women 1.83, 95% CI 1.72–1.95). Increased risk became evident at an early age (35–49 years) and was equally high or higher among women as compared with men. Lower socio-economic status and greater time living in Canada were also associated with increased risk for diabetes. Interpretation: Recent immigrants, particularly women and immigrants of South Asian and African origin, are at high risk for diabetes compared with long-term residents of Ontario. This risk becomes evident at an early age, suggesting that effective programs for prevention of diabetes should be developed and targeted to immigrants in all age groups.

Assessment of radiation dose awareness among pediatricians

BackgroundThere is increasing awareness among pediatric radiologists of the potential risks associated with ionizing radiation in medical imaging. However, it is not known whether there has been a corresponding increase in awareness among pediatricians.ObjectiveTo establish the level of awareness among pediatricians of the recent publicity on radiation risks in children, knowledge of the relative doses of radiological investigations, current practice regarding parent/patient discussions, and the sources of educational input.Materials and methodsMultiple-choice survey.ResultsOf 220 respondents, 105 (48%) were aware of the 2001 American Journal of Roentgenology articles on pediatric CT and radiation, though only 6% were correct in their estimate of the quoted lifetime excess cancer risk associated with radiation doses equivalent to pediatric CT. A sustained or transient increase in parent questioning regarding radiation doses had been noticed by 31%. When estimating the effective doses of various pediatric radiological investigations in chest radiograph (CXR) equivalents, 87% of all responses (and 94% of CT estimates) were underestimates. Only 15% of respondents were familiar with the ALARA principle. Only 14% of pediatricians recalled any relevant formal teaching during their specialty training. The survey response rate was 40%.ConclusionAwareness of radiation protection issues among pediatricians is generally low, with widespread underestimation of relative doses and risks.

Association of anti-smoking legislation with rates of hospital admission for cardiovascular and respiratory conditions

Background: Few studies have examined the impact of anti-smoking legislation on respiratory or cardiovascular conditions other than acute myocardial infarction. We studied rates of hospital admission attributable to three cardiovascular conditions (acute myocardial infarction, angina, and stroke) and three respiratory conditions (asthma, chronic obstructive pulmonary disease, and pneumonia or bronchitis) after the implementation of smoking bans. Methods: We calculated crude rates of admission to hospital in Toronto, Ontario, from January 1996 (three years before the first phase of a smoking ban was implemented) to March 2006 (two years after the last phase was implemented. We used an autoregressive integrated moving-average (ARIMA) model to test for a relation between smoking bans and admission rates. We compared our results with similar data from two Ontario municipalities that did not have smoking bans and with conditions (acute cholecystitis, bowel obstruction and appendicitis) that are not known to be related to second-hand smoke. Results: Crude rates of admission to hospital because of cardiovascular conditions decreased by 39% (95% CI 38%–40%) and admissions because of respiratory conditions decreased by 33% (95% CI 32%–34%) during the ban period affecting restaurant settings. No consistent reductions in these rates were evident after smoking bans affecting other settings. No significant reductions were observed in control cities or for control conditions. Interpretation: Our results serve to expand the list of health outcomes that may be ameliorated by smoking bans. Further research is needed to establish the types of settings in which smoking bans are most effective. Our results lend legitimacy to efforts to further reduce public exposure to tobacco smoke.

Immigrants’ duration of residence and adverse birth outcomes: a population‐based study

Please cite this paper as: Urquia M, Frank J, Moineddin R, Glazier R. Immigrants’ duration of residence and adverse birth outcomes: a population‐based study. BJOG 2010;117:591–601.

Correlation of sonographic findings and outcome in necrotizing enterocolitis

BackgroundThere is little in the literature regarding the use of gray-scale and Doppler sonography of the bowel in necrotizing enterocolitis (NEC) and how findings depicted by this modality might assist in predicting outcome and influence management.ObjectiveTo correlate sonographic findings with outcome in NEC.Materials and methodsThis was a retrospective analysis of clinical and abdominal ultrasonography (AUS) findings in NEC from January 2003 to December 2005. AUS findings were evaluated for portal venous gas, free gas, peritoneal fluid, bowel wall thickness, echogenicity, perfusion and intramural gas. Patients were categorized into two groups, according to their outcome.ResultsA total of 40 infants were identified who had AUS for NEC prior to any surgical intervention. Group A comprised 18 neonates treated medically and recovered fully, and group B comprised 22 neonates who required surgery or died. Free gas (six patients) and focal fluid collections (three patients) were only found in group B. Increased bowel wall echogenicity, absent bowel perfusion, portal venous gas, bowel wall thinning, bowel wall thickening, free fluid with echoes and intramural gas were seen in both groups, but more frequently in group B. Anechoic free fluid was seen more frequently in group A. Increased bowel perfusion was seen equally in both groups.ConclusionAn adverse outcome was associated with the sonographic findings of free gas, focal fluid collections or three or more of the following: increased bowel wall echogenicity, absent bowel perfusion, portal venous gas, bowel wall thinning, bowel wall thickening, free fluid with echoes and intramural gas. Sonographic findings are useful in predicting outcome and therefore might help guide management.

Population-Based Longitudinal Study of Follow-Up Care for Breast Cancer Survivors

PURPOSE To describe the patterns of follow-up care provided to a population-based cohort of breast cancer survivors, and to assess factors associated with adherence to guidelines on follow-up care. PATIENTS AND METHODS We conducted a retrospective longitudinal study of all women with surgically treated breast cancer who were without evidence of recurrence, advanced breast cancer, or new primary cancer and were diagnosed in Ontario, Canada, within a 2-year period (n = 11,219). They were followed for 5 years. The cohort was identified through the Ontario Cancer Registry, and individuals were linked across population-based administrative health databases. Frequency of and adherence to guideline recommendations for oncologist and primary care physician (PCP) visits; surveillance imaging for metastatic disease; and surveillance mammograms by year from diagnosis, age group, and income quintile were analyzed. Factors associated with adherence to guideline recommendations were analyzed. RESULTS Most women saw both oncologists and PCPs in each follow-up year. Approximately two thirds had surveillance mammograms in each follow-up year. Overall, two thirds had either fewer or greater than recommended oncology visits, one quarter had fewer than recommended surveillance mammograms, and half had greater than recommended surveillance imaging for metastatic disease. CONCLUSION This population-based study shows substantial variation in adherence to guideline recommendations, with both overuse and underuse of surveillance visits and tests. Most importantly, a substantial proportion are receiving more than recommended imaging for metastatic disease but fewer than recommended mammograms for detection of local recurrence or new primary cancer, for which effective intervention is possible.

Gabapentin for the treatment of menopausal hot flashes: a randomized controlled trial.

Objective: To compare the effectiveness and tolerability of gabapentin with placebo for the treatment of hot flashes in women who enter menopause naturally. Design: A randomized, double-blind, placebo-controlled trial was conducted across the greater Toronto area between March 2004 and April 2006 in the community and primary care settings. Eligible participants were 200 women in natural menopause, aged 45 to 65 years, having at least 14 hot flashes per week. Study participants were randomized to receive gabapentin 300 mg oral capsules or placebo three times daily for 4 weeks. The primary outcome measure was the mean percentage change from baseline to week 4 in daily hot flash score, determined from participant diaries. Secondary outcome measures included changes in weekly mean hot flash scores and frequencies, quality of life, and adverse events. Results: Of the 197 participants, 193 (98%) completed the study. Analysis was by intention to treat. Hot flash scores decreased by 51% (95% CI: 43%-58%) in the gabapentin group, compared with 26% (95% CI: 18%-35%) on placebo, from baseline to week 4. This twofold improvement was statistically significant (P < 0.001). The Menopause-Specific Quality-of-Life vasomotor score decreased by 1.7 (95% CI: 1.3-2.1; P < 0.001) in the gabapentin group. These women reported greater dizziness (18%), unsteadiness (14%), and drowsiness (12%) at week 1 compared with those taking placebo; however, these symptoms improved by week 2 and returned to baseline levels by week 4. Conclusions: Gabapentin at 900 mg/day is an effective and well-tolerated treatment for hot flashes.

Density, Destinations or Both? A Comparison of Measures of Walkability in Relation to Transportation Behaviors, Obesity and Diabetes in Toronto, Canada

The design of suburban communities encourages car dependency and discourages walking, characteristics that have been implicated in the rise of obesity. Walkability measures have been developed to capture these features of urban built environments. Our objective was to examine the individual and combined associations of residential density and the presence of walkable destinations, two of the most commonly used and potentially modifiable components of walkability measures, with transportation, overweight, obesity, and diabetes. We examined associations between a previously published walkability measure and transportation behaviors and health outcomes in Toronto, Canada, a city of 2.6 million people in 2011. Data sources included the Canada census, a transportation survey, a national health survey and a validated administrative diabetes database. We depicted interactions between residential density and the availability of walkable destinations graphically and examined them statistically using general linear modeling. Individuals living in more walkable areas were more than twice as likely to walk, bicycle or use public transit and were significantly less likely to drive or own a vehicle compared with those living in less walkable areas. Individuals in less walkable areas were up to one-third more likely to be obese or to have diabetes. Residential density and the availability of walkable destinations were each significantly associated with transportation and health outcomes. The combination of high levels of both measures was associated with the highest levels of walking or bicycling (p<0.0001) and public transit use (p<0.0026) and the lowest levels of automobile trips (p<0.0001), and diabetes prevalence (p<0.0001). We conclude that both residential density and the availability of walkable destinations are good measures of urban walkability and can be recommended for use by policy-makers, planners and public health officials. In our setting, the combination of both factors provided additional explanatory power.

Predictors of Non-Diagnostic Ultrasound Scanning in Children with Suspected Appendicitis

OBJECTIVE To determine predictors of diagnostically inaccurate ultrasound scanning for suspected appendicitis. STUDY DESIGN Prospective emergency department cohort study of 263 previously healthy children 4 to 17 years of age undergoing ultrasound scanning. Ultrasound scanning results were interpreted as positive, negative, or equivocal for appendicitis and classified as diagnostically accurate and inaccurate. The main outcome measure was association between inaccurate ultrasound scanning and age, sex, body mass index percentiles, pain duration, white blood cell count, Faces Pain Score-Revised, clinical probability of appendicitis, and ultrasound scanning operator. RESULTS Of the 263 patients, 95 ultrasound scanning examinations were read as positive, 76 as negative, and 92 were equivocal. A total of 162 (61.6%) ultrasound scanning examinations were accurate (TP86, TN76), and 101 (38.4%) ultrasound scanning examinations were inaccurate (FP88, FN13). Children with body mass index percentiles ≥ 85 and clinical probability of appendicitis ≤ 50% had 58.1% probability of inaccurate ultrasound scanning examination (odds ratio, 2.48; 95% confidence interval, 1.48-2.78). In lean children, diagnostic accuracy of the screening ultrasound scanning examination with second ultrasound scanning or clinical reassessment was 93% versus 83% in the obese children (95% confidence interval of the difference, 1-19%). CONCLUSION Screening ultrasound scanning for pediatric appendicitis has suboptimal accuracy, particularly in obese children with a low likelihood of appendicitis who should not routinely undergo ultrasound scanning. However, when followed by a second ultrasound scanning or a clinical reassessment, it offers high diagnostic accuracy in lean children.