PubMed Health⌕ Search

Biomedical subjects

J Sonis

Publications and source records attributed to J Sonis.

10 recordsLinked to original sources
PubMed1999

How to use and interpret interval likelihood ratios.

BACKGROUND: Likelihood ratios offer important advantages over sensitivity and specificity for characterizing diagnostic tests. They can capture the magnitude of abnormality of test results, whereas sensitivity and specificity require that the test results be dichotomized into positive or negative. This is an important advantage because many diagnostic tests are measured on continuous or ordinal scales. Posttest probabilities calculated from interval likelihood ratios may be different than those calculated from sensitivity and specificity; clinical decisions derived from the use of likelihood ratios may therefore be different from decisions derived from test results characterized by sensitivity and specificity. This article demonstrates the advantages, use, and interpretation of interval likelihood ratios using the clinical scenario of a young child with a high fever.

Age Factors↗
PubMed1998

Association between duration of residence and access to ambulatory care among Caribbean immigrant adolescents.

OBJECTIVES: The purpose of this study was to determine the association between duration of residence and access to ambulatory care among Caribbean immigrant adolescents. METHODS: A cross-sectional survey of adolescents at a New York City high school was conducted. Multivariate modeling methods were used to adjust for demographic and socioeconomic covariates. RESULTS: Duration of residence was strongly and directly associated with access to ambulatory care after adjustment for ethnicity. CONCLUSIONS: Caribbean immigrant adolescents, particularly recent immigrants, have reduced access to ambulatory care.

Acculturation↗
PubMed1998

A closer look at confounding.

BACKGROUND AND OBJECTIVES: Confounding is one of the most common and important biases in primary care research. This article explains the genesis and effects of two common misconceptions of confounding: 1) Confounding can be assessed with a statistical test. 2) All covariates should be included in a multivariate model to control confounding. Assessment of confounding by testing the statistical significance of baseline differences or the significance of a potential confounding factor in a multivariate model can produce underestimates or overestimates of the true association between an exposure and an outcome. Inclusion of all covariates in a multivariate model may lead to controlling for variables that are not, in fact, confounders. This may produce underestimates or overestimates of the effect in question, as well as artificially widened confidence intervals. Both of these misconceptions can lead to profound misinterpretation of research results. To prevent problems resulting from these misunderstandings, researchers should consider drawing causal models prior to conducting the research and use the change-in-estimate criterion, rather than a statistical test, to detect confounding.

Bias↗
PubMed1996

Teaching of human rights in US medical schools.

OBJECTIVE: To determine the extent to which human rights issues are included in required bioethics curricula in US medical schools and to identify medical school characteristics associated with the extent of human rights issues covered. DESIGN: Cross-sectional survey. PARTICIPANTS: Bioethics course directors and bioethics section directors of 125 US medical schools. MAIN OUTCOME MEASURE: The extent of human rights teaching at each school was measured as the percentage of 16 human rights issues. RESULTS: Course directors at 113 (90%) of the 125 US medical schools responded to the survey. Medical schools included about half (45%; 95% confidence interval, 41%-49%) of 16 human rights issues in their required bioethics curricula. Domestic human rights issues, such as discrimination in the provision of health care to minorities (82% of medical schools), were covered much more frequently than international human rights issues, such as physician participation in torture (17% of schools). Public medical schools included substantially fewer human rights issues than private medical schools (F[1,112]=7.7; P<.01). CONCLUSIONS: Required courses in medical education do not adequately address the medical aspects of human rights issues, especially international issues.

Analysis of Variance↗
PubMed1995

Family physicians and human rights: a case example from former Yugoslavia.

BACKGROUND: The role of family physicians in the protection of human rights is unclear. The purpose of this article is to explore this role through examination of the specific case of former Yugoslavia. METHODS: In January/February 1993, we investigated medical aspects of human rights abuses as part of a fact-finding mission for the Physicians for Human Rights organization. We used primarily qualitative methods, including direct observation, key informant interviews, focus groups, and key document review. RESULTS: We observed pervasive violations of medical neutrality, as indicated by the shelling of Kosevo Hospital, the major tertiary care hospital in Sarajevo. Forty-one percent of former prison camp detainees from north Bosnia housed at the Karlovac, Croatia, refugee camp had scars consistent with physical abuse. Civilians in north central Bosnia and Sarajevo were targeted through physical violence, artillery shelling, and strangulation of the public health infrastructure--food, water, electricity, heat, and medical supplies. CONCLUSIONS: Physicians can be uniquely useful in the investigation of human rights. The flexibility of family physicians and their attention to the biopsycosocial aspects of health can be especially useful. Family physicians should become actively involved in the protection of human rights.

Adult↗
PubMed1994

The quality of clinical trials published in The Journal of Family Practice, 1974-1991.

BACKGROUND: Previous analyses of published clinical trials have identified major deficiencies in reporting, design, analysis, and overall quality. The purpose of this study was to determine the strengths and weaknesses of published clinical trials in family practice, and to identify predictors of quality in these trials. METHODS: Randomized controlled clinical trials published in The Journal of Family Practice from 1974 to 1991 were eligible for the study. Two raters independently evaluated the adequacy and appropriateness of reporting, design, and analysis for each clinical trial, using the Chalmers index for assessing clinical trial quality. Multiple linear regression was used to determine the predictors of quality. RESULTS: The 53 trials included in the study showed deficiencies in reporting, design, and analysis, although fundamental design issues, such as blinding, were a relative strength. On average, the trials scored 35% of the possible points on the scale. Three factors were positively associated with overall quality: year of publication, number of pages of the published report, and the type of intervention. Trials with pharmacologic and non-medication therapy interventions, such as diet, had higher quality scores than did trials with psychosocial or educational interventions. CONCLUSIONS: The overall quality of these clinical trials was less than optimal but comparable to previously analyzed groups of trials. The improvement in quality over time may be related to improvement in the quality of the trials themselves, or more exacting editorial standards, or a combination of the two.

Bias↗