Open Reduction and Internal Fixation for Humeral Shaft Nonunion: Bone Grafting Is Not Routinely Required and Avoids Donor Site Morbidity

OBJECTIVESTo document union rate, complications and patient-reported outcomes after open reduction and internal fixation (ORIF), with and without bone grafting (BG), for humeral diaphyseal nonunion after failed nonoperative management. DESIGNRetrospective. SETTINGUniversity teaching hospital. PATIENTS AND INTERVENTIONFrom 2008 to 2017, 86 consecutive patients [mean age 59 years (range 17-86), 71% (n = 61/86) women] underwent nonunion ORIF (plate and screws) at a mean of 7 months postinjury (range 3-21.5). Eleven (13%) underwent supplementary BG. MAIN OUTCOME MEASUREMENTSUnion rate and complications for 83 patients (97%) at a mean of 10 months (3-61). Patient-reported outcomes (QuickDASH, EQ-5D, EQ-VAS, SF-12, satisfaction) for 53 living, cognitively-intact patients (78%) at a mean of 4.9 years (0.3-9.2). RESULTSNinety-three percent (n = 77/83) achieved union after nonunion ORIF. Complications included recalcitrant nonunion (7%, n = 6/83), iatrogenic radial nerve palsy (6%, n = 5/83), infection (superficial 7%, n = 6/83; deep 2%, n = 2/83), and iliac crest donor site morbidity (38%, n = 3/8). The union rate with BG was 78% (n = 7/9) and without was 95% (n = 70/74; P = 0.125), and was not associated with the nonunion type (atrophic 91%, n = 53/58; hypertrophic 96%, n = 24/25; P = 0.663). Median QuickDASH was 22.7 (0-95), EQ-5D 0.710 (-0.181-1), EQ-visual analog scale 80 (10-100), SF-12 physical component summary 41.9 (16-60.5), and mental component summary 52.6 (18.7-67.7). Nineteen percent (n = 10/53) were dissatisfied with their outcome. CONCLUSIONSORIF for humeral diaphyseal nonunion was associated with a high rate of union. Routine BG was not required and avoided the risk of donor site morbidity. One in 5 patients were dissatisfied despite the majority achieving union. LEVEL OF EVIDENCETherapeutic Level III. See Instructions for Authors for a complete description of levels of evidence.