Measures of Agreement for Dichotomous Outcomes

Measures of Agreement for Probabilistic Diagnoses

SSRN Electronic Journal ◽

10.2139/ssrn.2566937 ◽

2015 ◽

Author(s):

Douglas Mossman

Keyword(s):

Measures Of Agreement

Download Full-text

Ordinal outcome analysis improves the detection of between-hospital differences in outcome

BMC Medical Research Methodology ◽

10.1186/s12874-020-01185-7 ◽

2021 ◽

Vol 21 (1) ◽

Author(s):

I. E. Ceyisakar ◽

N. van Leeuwen ◽

Diederik W. J. Dippel ◽

Ewout W. Steyerberg ◽

H. F. Lingsma

Keyword(s):

Logistic Regression ◽

Quality Of Care ◽

Outcome Measures ◽

Statistical Power ◽

Ordinal Scale ◽

Efficiency Gain ◽

Number Of Patients ◽

Dichotomous Outcomes ◽

Ordinal Outcome

Abstract Background There is a growing interest in assessment of the quality of hospital care, based on outcome measures. Many quality of care comparisons rely on binary outcomes, for example mortality rates. Due to low numbers, the observed differences in outcome are partly subject to chance. We aimed to quantify the gain in efficiency by ordinal instead of binary outcome analyses for hospital comparisons. We analyzed patients with traumatic brain injury (TBI) and stroke as examples. Methods We sampled patients from two trials. We simulated ordinal and dichotomous outcomes based on the modified Rankin Scale (stroke) and Glasgow Outcome Scale (TBI) in scenarios with and without true differences between hospitals in outcome. The potential efficiency gain of ordinal outcomes, analyzed with ordinal logistic regression, compared to dichotomous outcomes, analyzed with binary logistic regression was expressed as the possible reduction in sample size while keeping the same statistical power to detect outliers. Results In the IMPACT study (9578 patients in 265 hospitals, mean number of patients per hospital = 36), the analysis of the ordinal scale rather than the dichotomized scale (‘unfavorable outcome’), allowed for up to 32% less patients in the analysis without a loss of power. In the PRACTISE trial (1657 patients in 12 hospitals, mean number of patients per hospital = 138), ordinal analysis allowed for 13% less patients. Compared to mortality, ordinal outcome analyses allowed for up to 37 to 63% less patients. Conclusions Ordinal analyses provide the statistical power of substantially larger studies which have been analyzed with dichotomization of endpoints. We advise to exploit ordinal outcome measures for hospital comparisons, in order to increase efficiency in quality of care measurements. Trial registration We do not report the results of a health care intervention.

Download Full-text

Timing of Pegfilgrastim: Association with Febrile Neutropenia in a Pediatric Solid and CNS Tumor Population

Journal of Pediatric Oncology Nursing ◽

10.1177/10434542211037729 ◽

2021 ◽

Vol 38 (6) ◽

pp. 375-384

Author(s):

Laura Schlenker ◽

Renee C.B. Manworren

Keyword(s):

Febrile Neutropenia ◽

Mixed Model ◽

Random Effect ◽

Retrospective Chart Review ◽

Patient Characteristics ◽

Secondary Outcome ◽

Pharmacokinetic Data ◽

Risk Patients ◽

Independent Variable ◽

Dichotomous Outcomes

Background: While recommended timing of pegfilgrastim administration is ≥24 h after chemotherapy, patient barriers to next day administration, available adult evidence, and pharmacokinetic data have led to earlier administration in some pediatric patients with solid and central nervous system tumors. The purpose of this study was to compare patient outcomes by timing of pegfilgrastim after chemotherapy. Methods: A retrospective chart review examined timing of 932 pegfilgrastim administrations to 182 patients, 0–29 years of age. The primary outcome was febrile neutropenia (FN); the secondary outcome was neutropenic delays (ND) ≥7 days to next chemotherapy cycle. To account for multiple pegfilgrastim administrations per patient, a generalized mixed model was used with a logit link for the dichotomous outcomes (FN & ND), timing as the dichotomous independent variable, and random effect for patient. Results: FN occurred in 196 of 916 cycles (21.4%); and ND in 19 of 805 cycles (2.4%). The fixed effect of pegfilgrastim administration < or ≥24 h after chemotherapy was not significant, p = .50; however, earlier or later than 20 h was significant, p = .005. FN odds were significantly higher when pegfilgrastim was given <20 h (OR 1.78, 95% CI: 1.19–2.65) after chemotherapy, which may be attributable to differences in chemotherapy toxicity regardless of pegfilgrastim timing. Discussion: While attempts should be made to administer pegfilgrastim ≥24 h after chemotherapy, if barriers exist, modified timing based on individual patient characteristics should be considered. Prospective randomized trials are needed to identify lower risk patients for early pegfilgrastim administration.

Download Full-text