gms | German Medical Science

Background: Acute respiratory distress syndrome (ARDS)/acute lung injury (ALI) is a life-threatening condition resulting from inflammatory-mediated damage of lung tissue. Leading causes of ARDS are pneumonia, trauma, sepsis and aspiration and in-hospital mortality is high (>30%). ARDS outcomes research has largely focused on mortality, but has neglected longer-term outcomes such as health related quality of life (HRQoL) and long-term disability in survivors of ARDS.

Research Questions: Which factors determine HRQoL and return to work (RtW) in survivors of ARDS?

Methods: A systematic literature review was conducted. We searched PubMed, Embase, PsycInfo, and Web of Science. The following eligibility criteria were applied. Naturalistic observational studies (cohort, case-control and cross-sectional studies) in ARDS/ALI survivors which reported (a) statistical association(s) between (a) determinant(s) of any kind and HRQoL and/ or RtW were eligible for inclusion. Study eligibility and risk of bias were independently assessed by four researchers (working in two pairs). Data extraction included the characteristics of the study (country, study type etc.) and of the sample/ cohort (N, age and sex). Additionally, information on the outcome measurement instruments for HRQoL and RTW and the analyzed determinants (p-values and effects) were extracted. Risk of bias was assessed with respect to selection and information bias.

Results: 4559 records were retrieved from electronic searches, of which 24 studies were included in this systematic review. Sociodemographic determinants (sex and age) were investigated by six studies, whereby analyzes regarding age were most frequent. Across all studies old age was associated with lower HRQoL. Associations with RtW were not reported at all. In the section of disease related determinants the results of lung function testing were often correlated with the scores of the HRQoL assessment. Even the performance of lung functioning was measured in various ways, positive associations between HRQoL/RtW and lung function predominated. The use of extracorporeal lung assist techniques, which was the most frequently investigated determinant in the section of care related factors, resulted in no or negative associations with HRQoL/RtW. In the section of psychosocial determinants depression and posttraumatic stress disorder were negative predictors for HRQoL and RtW. Overall, high consistency of effects across determinants investigated in more than one study was obtained. Risk of information bias was low in nearly all of the studies. This results from the use of validated measurement instruments or secured records for the assessment of the determinants and HRQoL. Solely the approach used in the assessment of RtW remained unclear in some studies. A high risk of selection bias was immanent in 11 studies. In these cases the selection of the sample/ cohort remained unclear or no attempt was made to draw a random sample or to include the complete population of interest.

Discussion: This systematic review revealed a broad range of determinants of HRQoL in ARDS/ALI survivors in the respective research literature. Determinants were found in all four domains (sociodemographic, disease-related, care-related and psychosocial). Supportive measures like extracorporeal lung assist techniques seem to have a negative association with HQRoL and RtW. An obvious reason for this could be the absence of statistical adjustment for disease severity. While most studies used validated measurement instruments and thus had only low risk of information bias, risk of selection bias was high in some studies. The recruitment and the (long-term) follow-up of patients surviving a critical illness such as ARDS seem challenging and warrant high efforts. The observational design of the studies requires an appropriate statistical analysis, which takes potential covariates into account. However, in most studies this demand was not met.

Implications: There is a lack of high quality observational studies investigating determinants of HRQol and RtW in survivors of ARDS. The risk of bias in most of the existing studies is substantial, limiting both internal and external validity. The evaluation of potentially modifiable care-related factors is thus far underrepresented.