Non-union Rates in Hind and Midfoot Arthrodesis in Current, Ex-, and Non-smokers

Abstract

Category: Hindfoot Introduction/Purpose: Arthrodesis is a safe and effective treatment for a number of hind and midfoot conditions. However, non-union rates have historically been reported as high as 41%. A number of factors have been identified that increase non-union rates, the most notable and readily modifiable is a patient’s smoking status. Smoking rates in the UK in 2015 were 19.3% for men and 15.3% for women. We have examined the effect of smoking status (current, ex-smokers and non-smokers) on union rates for a large cohort of patients undergoing hind or midfoot arthrodesis. Methods: This is a single surgeon, retrospective cohort study of consecutive cases. The surgeon’s logbook was used to identify patients undergoing any hind and midfoot arthrodesis procedures from January 2010 until September 2016. Revision procedures and charcot arthropathy cases were excluded along with cases with insufficient records available. Demographic data was collected along with: joints involved, surgical implant used, bone grafting, the use of ultrasound bone stimulation (EXOGEN, Bioventus LLC, Durham, USA) therapy, complications and final outcome with regards to union. Patients were divided according to self-reported smoking status at pre-operative assessment; current smokers, ex-smokers and non-smokers. Union outcome was based on clinical notes and included patient symptoms and radiographic evidence. Delayed union was classed as union occurring after 6 months. The effect of smoking status on deep infection rates and the need for EXOGEN therapy was also analysed. Results: 381 joints were included (see image). The smoking prevalence was 14.0% (accounting for 12.3% of joints) and 32.2% ex-smokers (35.4%). The groups were comparable with regards to gender, diabetes status and BMI. Smokers were younger, had less co-morbidities and were less likely to have had multiple joints fused (p<0.05). Non-union rates were statistically higher in current smokers with a relative risk of 5.81 (95% CI 2.54-13.29, P<0.001), there was no statistically significant difference between ex-smokers and non-smokers. Active smokers had higher rates of deep infection (P=0.05) and the need for EXOGEN use (P<0.001). Within the smoking group there was a trend toward slower union (delayed + non-union) with heavier smoking (p=0.054). Conclusion: This large retrospective cohort study confirms previous evidence that smoking has a considerable negative effect on union in arthrodesis (despite other differences between the groups likely to favour union in smokers). A 5.8 relative risk in a modifiable risk factor is extremely high. Arthrodesis surgery should be undertaken with extreme caution in patients who are actively smoking. Our study shows that after cessation of smoking the risk returns to normal, however we have not quantified the time frame. Further research is needed to quantify the necessary time frame for smoking cessation to reduce non-union risk.