A comparison of outcome measures used to report clubfoot treatment with the Ponseti method: results from a cohort in Harare, Zimbabwe.

BACKGROUND: There are various established scoring systems to assess the outcome of clubfoot treatment after correction with the Ponseti method. We used five measures to compare the results in a cohort of children followed up for between 3.5 to 5 years. METHODS: In January 2017 two experienced physiotherapists assessed children who had started treatment between 2011 and 2013 in one clinic in Harare, Zimbabwe. The length of time in treatment was documented. The Roye score, Bangla clubfoot assessment tool, the Assessing Clubfoot Treatment (ACT) tool, proportion of relapsed and of plantigrade feet were used to assess the outcome of treatment in the cohort. Inter-observer variation was calculated for the two physiotherapists. A comparative analysis of the entire cohort, the children who had completed casting and the children who completed more than two years of bracing was undertaken. Diagnostic accuracy was calculated for the five measures and compared to full clinical assessment (gold standard) and whether referral for further intervention was required for re-casting or surgical review. RESULTS: 31% (68/218) of the cohort attended for examination and were assessed. Of the children who were assessed, 24 (35%) had attended clinic reviews for 4-5 years, and 30 (44%) for less than 2 years. There was good inter-observer agreement between the two expert physiotherapists on all assessment tools. Overall success of treatment varied between 56 and 93% using the different outcome measures. The relapse assessment had the highest unnecessary referrals (19.1%), and the Roye score the highest proportion of missed referrals (22.7%). The ACT and Bangla score missed the fewest number of referrals (7.4%). The Bangla score demonstrated 79.2% (95%CI: 57.8-92.9%) sensitivity and 79.5% (95%CI: 64.7-90.2%) specificity and the ACT score had 79.2% (95%CI: 57.8-92.9%) sensitivity and 100% (95%CI: 92-100%) specificity in predicting the need for referral. CONCLUSION: At three to five years of follow up, the Ponseti method has a good success rate that improves if the child has completed casting and at least two years of bracing. The ACT score demonstrates good diagnostic accuracy for the need for referral for further intervention (specialist opinion or further casting). All tools demonstrated good reliability