Effectiveness of telephone prompts when surveying general practitioners: a randomised trial

Abstract

Background: Response rates to recent surveys of general practitioners have ranged from 44-95%. The effectiveness of a telephone prompt by a non-medical research assistant in advance of survey mail-out was unknown at the time of this study.

Objective: To determine the effectiveness of a telephone prompt by a non-medical research assistant to enhance response rates to a lengthy survey.

Subjects: 364 randomly selected general practitioners randomised into intervention and control groups.

Outcome variable: Cumulative response rate at Day 18 and Day 60.

Design: Intervention group received a telephone call in advance of the survey. The control group received conventional mail-outs. Non-responders in both groups were telephoned at Day 28 and sent a reminder letter and questionnaire.

Results: 77% of the intervention group responded to the survey compared to 64% of the control group. Intention-to-treat analysis demonstrated a significantly earlier response by Day 18 (continuity corrected chi 2 = 18.15; df = 1; p < 0.001) and a significantly higher overall response by Day 60 (continuity corrected chi 2 = 5.41, df = 1, p < 0.009) from the intervention group.

Conclusion: A telephone prompt by a non-medical research assistant will accelerate and enhance response rates to a survey of general practitioners although the differential effectiveness of a non-medical researcher compared to a medical practitioner conducting such prompts remains unstudied.