Limiting loss to follow-up in a multicenter randomized trial in orthopedic surgery

Even the best-designed, randomized controlled trials suffer when patients are lost to follow-up. Incomplete follow-up biases the results of a trial when patients who drop out are different from those who complete follow-up. This is exaggerated further when there are differential dropout rates between study groups. Previous randomized controlled trials in orthopedic trauma have reported up to 28% loss to follow-up. Only by striving to achieve a 0% loss to follow-up rate can we be certain that this type of bias does not affect our results. In our ongoing multicenter, randomized controlled trial comparing reamed and nonreamed intramedullary nailing of tibial shaft fractures, we have implemented several innovative strategies to minimize loss to follow-up. The exclusion criteria and consent process are designed to minimize losses. Study staff are carefully trained in communication and negotiation with patients. Additionally, a central methods center monitors all patient follow-up and aids in finding lost patients. Through these primary, secondary, and tertiary interventions, we have achieved 94% complete 1-year follow-up for the first 440 patients enrolled in the trial. Eleven patients withdrew consent, and we are unable to locate 17 patients. We have successfully minimized the loss to follow-up rate in our trial by incorporating innovative prevention and retention strategies into its design and conduct. Through planning, organization, and committing time and resources to minimizing loss to follow-up, other orthopedic trauma trials can hope to achieve the same high rates of follow-up.