Extending the PICOTS Framework for Evidence Syntheses of Digital Health Interventions: Results of a Delphi Consensus Study

Author(s)

Zrubka Z1, Champion A2, Holtorf AP3, Di Bidino R4, Earla JR5, Boltyenkov A6, Tabata-Kelly M7, Asche C8, Burrell A9
1Óbuda University, Budapest, Hungary, 2Healthcare Research Insights, Inc, Lake Forest, IL, USA, 3Health Outcomes Strategies GmbH, Basel, Switzerland, 4A.Gemelli Teaching Hospital, Roma, Italy, 5Merck, Rahway, NJ, USA, 6Siemens Healthcare Diagnostics Inc., Hellertown, PA, USA, 7Brandeis University, Waltham, MA, USA, 8University of Utah, Salt Lake City, UT, USA, 9Anita Burrell Consulting LLC, Flemington, NJ, USA

OBJECTIVES: Previous research has shown that definitions of digital health-related terms are diverse and ambiguous. Therefore, we aimed to develop a minimum information framework for defining research questions for patient-facing digital health interventions that is sufficiently specific for evidence syntheses in health economics and outcomes research (HEOR).

METHODS: Building on a comprehensive review and qualitative content analysis of digital health- related definitions in systematic reviews, the ISPOR Digital Health Special Interest Group (DH-SIG) launched a Delphi survey study to reach consensus on items of a minimum information framework, following the CREDES guidelines. Altogether, 28 proposed subcategories were organized under the domains of PICOTS (“Population”, “Intervention”, “Comparator”, “Outcome”, “Timing” and “Setting”) extended with “Information”, “Communication”, and “Technology”. Volunteering members of the DH-SIG with at least moderate familiarity with digital health were invited to rate, comment and modify the proposed items in the first round of the Delphi survey, followed by two rounds of repeat ratings and feedback on the aggregated results. Consensus was predefined as ³70% “important” or “very important” ratings in the final round.

RESULTS: After screening 45 DH-SIG-members, 31 were eligible for the Delphi survey. The first, second and third rounds were completed partially or fully by 18, 11 and 10 respondents, respectively. The panel (18 first-round respondents) was diverse in terms of age, gender, geography, and professional background. Following panelists’ feedback, the domain “Context” and five subcategories were added in round two. In the third round, all 33 subcategories and nine of the ten domains reached consensus for inclusion in the minimum information set; with the “Information” domain excluded.

CONCLUSIONS: When defining PICOTS research questions on patient-facing digital health interventions in addition to the population, intervention, comparator, outcome, timing and setting, details related to communication, technology and context should be specified to aid evidence syntheses for HEOR purposes.