Dados do Trabalho

Título

METHODOLOGICAL QUALITY AND OUTCOME ASSESSMENT ON RANDOMIZED CLINICAL TRIALS FOR THE TREATMENT OF INSOMNIA

Introdução

The number of randomized controlled trials (RCTs) evaluating interventions for the treatment of insomnia has been increasing. However, the methodological quality and outcome assessment methods used in these RCTs vary considerably.

Objetivo

To describe the quality of RCTs, and the outcomes used to access insomnia.

Métodos

This is a secondary analysis of the systematic review used in the 2023 Brazilian Guidelines on the Diagnosis and Treatment of Insomnia in Adults, which included RCTs evaluating therapeutic interventions for non-comorbid insomnia. All records were evaluated by 2 reviewers in a 2-steps process (abstracts and full texts). Methodological quality was evaluated using the van Tulder scale, a 11 items tool evaluating common biases on RCTs. Sleep-related outcomes were analyzed by quantifying the proportion of included manuscript reporting the most used sleep assessment tools (including polysomnography, sleep logs, and sleep-related questionnaires). Non-sleep-related outcomes were analyzed using the taxonomy for outcome classification proposed by the Core Outcome Measures in Effectiveness Trials (COMET).

Resultados

132 RCTs were included, 58 related to pharmacological interventions, 71 to non-pharmacological ones, and 3 to both interventions. The average compliance rate was 67.4±29.9% among all the included articles, being significantly higher in those evaluating pharmacological interventions (77.7±12.5%) in comparison to non-pharmacological ones (59.3±13.6%, p<0.001). The higher compliance rates were related to “similar timing of outcome assessment” (96.7%, n=128), “similar baseline characteristics” (96.7%, n=128), and “absence of cointerventions” (96.2%, n=127). Lower compliance rates were observed on “acceptable compliance in all groups” (24.2%, n=32) and “intention-to-treat analyses” (25.0%, n=33). Non-pharmacological trials were associated with lower methodological quality in 5 items in the van Tulder scale (p<0.05). The most often reported outcome for insomnia was Insomnia Severity Index (49.2%, n=65), followed by sleep log measures (47.7%, n=63) and polysomnographic outcomes (31.1%, n=41). Regarding the non-sleep outcomes, the most often investigated were related emotional functioning/well-being (37.9%, n=50), followed by cognitive functioning (14.4%, n=19).

Conclusões

While the average quality of the included RCTs is acceptable, the average methodological quality is higher among studies evaluating pharmacological interventions.

Palavras -chave

Sleep; meta-epidemiology; systematic reviews; hypnotics.