Understanding generalisability, and issues of recruitment in cohort studies

: the example of advanced CKD in the elderly

Abstract

Introduction: Observational studies play a valuable role in nephrology and have informed clinical practice. A significant challenge to observational research is external validity. Threats to generalisability can occur at the design stage due to selective inclusion criteria, at the recruitment stage due to non-participation of specific groups, and finally at the reporting stage due to poor reporting.

Methods: The PhD thesis was a mixed method study of convergent parallel/ triangulation design, embedded into European QUALity Study on treatment in advanced chronic kidney disease (EQUAL). The thesis involved three studies: quantitative, qualitative and systematic review to understand more fully the various factors that could affect a study’s generalisability at the design, recruitment and reporting stages respectively. The quantitative arm was a retrospective observational study comparing patients in primary and secondary care meeting the same inclusion criteria as EQUAL. The qualitative arm involved semi-structured interviews with patients who agreed and did not agree to participate in EQUAL. The systematic review assessed the quality of reporting of cohort studies before (1/1/2002-31/12/2007) and after (1/10/2008-31/12/2013) the publication of the STROBE statement.

Results: The quantitative arm of the thesis showed that patients in EQUAL were more likely to be younger, male and from an urban setting compared to the primary and secondary care cohort patients. Overall lesser co-morbidity of EQUAL patients meant that they were more likely to be alive at one year or to be admitted to hospital for illnesses. In the qualitative arm of the thesis, patients who agreed to participate in research reported being activated in their healthcare, and this seemed to relate to their decision to take part in research. Altruistic morals had a strong influence on participation in EQUAL study. The issue of caring responsibility and transportation were the main reasons causing inconvenience and negatively influenced participation. The systematic review showed that there had been an improvement in the overall reporting quality of CKD cohort studies particularly in the latter three years of the post-STROBE period.

Conclusion: The mixed methods approach of this PhD thesis aided a breadth and depth of understanding of some of the issues affecting generalisability with corroboration of the results from the quantitative and qualitative arms of the study. Sustained efforts of researchers are required to improve the generalisability of research findings at every stage of observational research.

Date of Award	25 Jun 2019
Original language	English
Awarding Institution	The University of Bristol
Sponsors	UK Renal Registry
Supervisor	Lucy A Biddle (Supervisor), Fergus J Caskey (Supervisor) & Yoav Ben-Shlomo (Supervisor)