Objective: To determine the non-adherence to the primary care 'do not do' recommendations (DNDs) and their likelihood to cause harm.
Design: Delphi study.
Setting: Spanish National Health System.
Participants: A total of 128 professionals were recruited (50 general practitioners [GPs], 28 pediatricians [PEDs], 31 nurses who care for adult patients [RNs] and 19 pediatric nurses [PNs]).
Interventions: A selection of 27 DNDs directed at GPs, 8 at PEDs, 9 at RNs and 4 at PNs were included in the Delphi technique. A 10-point scale was used to assess whether a given practice was still present and the likelihood of it causing of an adverse event.
Main outcome measure: Impact calculated by multiplying an event's frequency and likelihood to cause harm.
Results: A total of 100 professionals responded to wave 1 (78% response rate) and 97 of them to wave 2 (97% response rate). In all, 22% (6/27) of the practices for GPs, 12% (1/8) for PEDs, 33% (3/9) for RNs and none for PNs were cataloged as frequent. A total of 37% (10/27) of these practices for GPs, 25% (2/8) for PEDs, 33% (3/9) for RNs and 25% (1/4) for PNs were considered as potential causes of harm. Only 26% (7/27) of the DNDs for GPs showed scores equal to or higher than 36 points. The impact measure was higher for ordering benzodiazepines to treat insomnia, agitation or delirium in elderly patients (mean = 57.8, SD = 25.3).
Conclusions: Low-value and potentially dangerous practices were identified; avoiding these could improve care quality.
Keywords: Delphi technique; overuse; patient safety; primary care; quality assurance.
© The Author(s) 2018. Published by Oxford University Press in association with the International Society for Quality in Health Care. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.