Enrolling patients without capacity to trauma trials; successes and challenges
WALLIS S., LERNER R., ACHTEN J., GRIFFIN X., COSTA M., TUTTON E.
Introduction: Patients lacking capacity are frequently excluded from research populations. The results of research carried out without these patients, may not be generalisable and patients lacking capacity could miss out on access to evidence-based treatment. The reasons for excluding patients based on capacity vary, but it has been suggested that recruiting and collecting outcomes is difficult, and that researchers are unsure of the ethical implications. Lack of capacity has been reported in 30- 40% in hip fracture patients. The World Hip Trauma Evaluation (WHiTE) trials, investigating this patient group, are amongst the largest studies to include patients without capacity. Aim: This paper describes the process used by WHiTE to recruit patients lacking capacity, the retention of patients and success of follow-up. Method: Patients lacking capacity are usually enrolled by a professional consultee (PrC) before surgery. After surgery, if appropriate, patients will either be asked to consent to continuing participation or a personal consultee (PeC) will be approached. We recorded data on pre- and post-enrollment consent types, withdrawals, and 120-day retention rate. Results: 2,146 patients enrolled between June 2016 and April 2019 - 1,461 (68%) were enrolled using consultee agreement. After surgery, of the 1,384 patients enrolled by PrC, 25% continued under patient consent, 36% under PeC agreement and 18% under continued PrC agreement. Only 7% of patients/PeC declined to continue with the study. 71% of patients initially enrolled by Consultees provided 120-day follow-up, compared to 89% of those enrolled under patient consent. Conclusion: Recruiting and collecting data from a frail patient group lacking capacity is feasible. The vast majority of participants who regained capacity agreed to continue participating. Whilst retention rates for patients who initially lacked capacity were slightly lower than those for patients who provided consent, inclusion of this data increased the overall generalisability of the data.