Ellen Kilsdonk

Factors influencing implementation success of guideline-based clinical decision support systems: A systematic review and gaps analysis

OBJECTIVE To provide an integrated and differentiated understanding of factors influencing guideline-based CDSS implementation and illustrate the gaps in the current literature. MATERIALS AND METHODS A systematic literature review and in-depth exploration of factors impeding or facilitating successful implementation of guideline-based CDSS supporting physicians in clinical decision-making was performed. Factors were identified thematically by textual analysis of the included publications and were individually mapped to the human, organization and technology-fit (HOT-fit) framework for evaluating implementations of health information systems. RESULTS A total of 421 factors were found in 35 included publications from a total of 3676 publications. The mapping of factors concerning CDSS implementation on the HOT-fit framework revealed gaps in each domain of the framework and showed that research has mainly focused on human and technology factors and less on organizational factors. CONCLUSIONS Future research within the field of guideline-based CDSS should focus on evaluating implementations through the use of socio-technical models to study guideline-based CDSS system implementations from a multidimensional view. Furthermore, research is needed to explore whether use of these models during the planning phases of a CDSS project is useful in anticipating and preventing implementation barriers from occurring and exploiting facilitators to a successful implementation of the system.

Participation rates of childhood cancer survivors to self-administered questionnaires: a systematic review

&NA; This review aimed to assess participation rates of childhood cancer survivors (CCS) invited to fill out a health‐related questionnaire. Additionally, effects of study and CCS characteristics on participation rates were examined. PubMed, Web of Science, Ovid (EMBASE) and CINAHL databases were searched. Publications included were questionnaire‐based studies among CCS diagnosed with cancer before the age of 21, alive at least 5 years past diagnosis and aged 16 years or older at the time of study. Thirty‐five studies were included; the median participation rate was 65%. Sixteen studies reported information about CCS actively declining participation (median rate 5%). Five study characteristics seemed to influence participation rates: the use of reminders and incentives, the option to answer a shortened questionnaire, the recruitment of participants through their general practitioner and a pre‐notification before sending out the questionnaire. Furthermore, CCS characteristics related to improved participation were female gender, Caucasian ethnicity and a higher educational level. The results of this study will help to improve the (methodological) quality of future questionnaire‐based studies among CCS, thereby increasing our knowledge about late effects among this group of survivors.

Effect of Web-Based Versus Paper-Based Questionnaires and Follow-Up Strategies on Participation Rates of Dutch Childhood Cancer Survivors: A Randomized Controlled Trial

Background Questionnaires are widely used in survey research, especially in cohort studies. However, participation in questionnaire studies has been declining over the past decades. Because high participation rates are needed to limit the risk of selection bias and produce valid results, it is important to investigate invitation strategies which may improve participation. Objectives The purpose of this study is to investigate the effect of Web-based versus paper-based questionnaires on participation rates in a questionnaire survey on late effects among childhood cancer survivors (CCSs). Methods A total of 750 CCSs were randomized across 3 study arms. The initial invitation in study arms 1 and 2 consisted of a Web-based questionnaire only, whereas in study arm 3 this invitation was complemented with a paper-based version of the questionnaire. The first postal reminder, sent to the nonresponding CCSs in all 3 study arms, consisted of either a reminder letter only (study arms 1 and 3) or a reminder letter complemented with a paper-based questionnaire (study arm 2). The second postal reminder was restricted to CCSs in study arms 1 and 2, with only those in study arm 1 also receiving a paper-based questionnaire. CCSs in study arm 3 received a second reminder by telephone instead of by mail. In contrast to CCSs in study arm 3, CCSs in study arms 1 and 2 received a third reminder, this time by telephone. Results: Overall, 58.1% (436/750) of the CCSs participated in the survey. Participation rates were equal in all 3 study arms with 57.4% (143/249) in arm 1, 60.6% (152/251) in arm 2, and 56.4% (141/250) in arm 3 (P=.09). Participation rates of CCSs who received an initial invitation for the Web-based questionnaire only and CCSs who received an invitation to complete either a paper-based or Web-based questionnaire did not differ (P=.55). After the first postal reminder, participation rates of CCSs invited for the Web-based questionnaire only also did not differ compared with CCSs invited for both the Web-based and paper-based questionnaires (P=.48). In general, CCSs preferred the paper-based over the Web-based questionnaire, and those completing the paper-based questionnaire were more often unemployed (P=.004) and lower educated (P<.001). Conclusion Invitation strategies offering a Web-based questionnaire without a paper-based alternative at first invitation can be used without compromising participation rates of CCS. Offering the choice between paper- and Web-based questionnaires seems to result in the highest accrual participation rate. Future research should look into the quality of the data delivered by both questionnaires filled in by respondents themselves. Trial Registration International Standard Randomized Controlled Trial Number (ISRCTN): 84711754; http://www.controlled-trials.com/ISRCTN84711754 (Archived by WebCite at http://www.webcitation.org/6c9ZB8paX)