Matching the Outcomes to Treatment Targets of Exercise for Low Back Pain: Does it Make a Difference? Results of Secondary Analyses From Individual Patient Data of Randomised Controlled Trials and Pooling of Results Across Trials in Comparative Meta-analysis.

OBJECTIVE: To explore whether using a single matched or composite outcome might affect the results of previous randomized controlled trials (RCTs) testing exercise for non-specific low back pain (NSLBP). The first objective was to explore whether a single matched outcome generated greater standardized mean differences (SMDs) when compared with the original unmatched primary outcome SMD. The second objective was to explore whether a composite measure, composed of matched outcomes, generated a greater SMD when compared with the original primary outcome SMD. DESIGN: We conducted exploratory secondary analyses of data. SETTING: Seven RCTs were included, of which 2 were based in the USA (University research clinic, Veterans Affairs medical center) and the UK (primary care clinics, nonmedical centers). One each were based in Norway (clinics), Brazil (primary care), and Japan (outpatient clinics). PARTICIPANTS: The first analysis comprised 1) 5 RCTs (n=1033) that used an unmatched primary outcome but included (some) matched outcomes as secondary outcomes, and the second analysis comprised 2) 4 RCTs (n=864) that included multiple matched outcomes by developing composite outcomes (N=1897). INTERVENTION: Exercise compared with no exercise. MAIN OUTCOME MEASURES: The composite consisted of standardized averaged matched outcomes. All analyses replicated the RCTs' primary outcome analyses. RESULTS: Of 5 RCTs, 3 had greater SMDs with matched outcomes (pooled effect SMD 0.30 [95% confidence interval {CI} 0.04, 0.56], P=.02) compared with an unmatched primary outcome (pooled effect SMD 0.19 [95% CI -0.03, 0.40] P=.09). Of 4 composite outcome analyses, 3 RCTs had greater SMDs in the composite outcome (pooled effect SMD 0.28 [95% CI 0.05, 0.51] P=.02) compared with the primary outcome (pooled effect SMD 0.24 [95% CI -0.04, 0.53] P=.10). CONCLUSIONS: These exploratory analyses suggest that using an outcome matched to exercise treatment targets in NSLBP RCTs may produce greater SMDs than an unmatched primary outcome. Composite outcomes could offer a meaningful way of investigating superiority of exercise than single domain outcomes.

Análisis Secundario de Datos Poblacionales en Salud: Consideraciones sobre el Uso de Encuestas / Secondary Analysis of Population Health Data: Considerations on the Use of Surveys

Introducción. El análisis de datos secundarios en salud es importante para el éxito de la salud de la población. El propósito de esta guía práctica es entregar recomendaciones para el análisis adecuado de datos secundarios en salud. Desarrollo. varios problemas asociados con el análisis secundario de los datos de encuestas de salud deben abordarse, estos se relacionan con el diseño de la muestra, la medición de los datos, la falta de respuesta y la pérdida de datos. Para abordar algunos de estos problemas, se sugiere considerar el peso de la muestra, la operacionalización de las variables y la imputación de datos. Discusión. el análisis inapropiado de datos puede dar lugar a conclusiones inexactas y afectar la confiabilidad y validez de evidencia generada. Es importante explorar el conjunto de datos a analizar. Conclusión. es importante ser consciente de las particularidades del análisis secundario para evitar errores previsibles al seleccionar un conjunto de datos y realizar análisis estadístico.

Background. The analysis of population health data is important for the success of population health. The purpose of this practical guide is to provide recommendations for the adequate analysis of secondary health data. Development. There are several issues associated with the secondary analysis of health survey data that need to be addressed, these include relational sample design, data collection, non-response, and missing data. To resolve some of these problems, it is suggested to consider the weight of the sample, the operationalization of the variables and the imputation of data. Discussion. The inappropriate analysis of data can lead to inaccurate conclusions and affect the reliability and validity of evidence produced. It is important explore the data set to analyze. Conclusion. It is important to be aware of the particularities of secondary analysis to avoid foreseeable errors when selecting a data set and performing statistical analysis.

Population differences and the effect of vaginal progesterone on preterm birth in women with threatened preterm labor (.).

OBJECTIVE: Threatened preterm labor (tPTL) is a complication of pregnancy. Identification of women and clinical definition differs between countries. This study investigated differences in tPTL and effectiveness of vaginal progesterone to prevent preterm birth (PTB) between two countries. METHODS: Secondary analysis of a randomized controlled trial (RCT) from Argentina and Switzerland comparing vaginal progesterone to placebo in women with tPTL (n = 379). Cox proportional hazards analysis was performed to compare placebo groups of both countries and to compare progesterone to placebo within each country. We adjusted for baseline differences. Iatrogenic onset of labor or pregnancy beyond gestational age of interest was censored. RESULTS: Swiss and Argentinian women were different on baseline. Risks for delivery <14 days and PTB < 34 and < 37 weeks were increased in Argentina compared to Switzerland, HR 3.3 (95% CI 0.62-18), 54 (95% CI 5.1-569) and 3.1 (95% CI 1.1-8.4). In Switzerland, progesterone increased the risk for delivery <14 days [HR 4.4 (95% CI 1.3-15.7)] and PTB <37 weeks [HR 2.5 (95% CI 1.4-4.8)], in Argentina there was no such effect. CONCLUSION: In women with tPTL, the effect of progesterone may vary due to population differences. Differences in populations should be considered in multicenter RCTs.