Incidence of Readmission Following Pediatric Hand Surgery: An Analysis of 6600 Patients

There has been increasing pressure on health care entities to demonstrate quality of care. The 30-day readmission rate following a surgical procedure has been identified as an important surrogate for health care quality. Consequently, understanding risks associated with readmission is relevant for patient counseling, risk stratification, and risk reduction. The Centers for Medicare & Medicaid services implemented a Hospital Readmissions Reduction Program in 2012, assigning a ratio to “standardize” unplanned readmissions and compare hospitals using this metric.¹ Whereas adult hand surgery has been associated with a low readmission rate, few studies address the risk factors for readmission in the pediatric population.^2-7

The American College of Surgeons (ACS) National Surgical Quality Improvement Program – Pediatric (NSQIP-P) maintains a prospective database from 50 children’s hospitals in the United States. These databases have been used to assess 30-day complications and readmissions in multiple disciplines within pediatric surgery.^8-14 An analysis of the NSQIP-P database is able to provide more complete data regarding readmissions following pediatric hand procedures.

This study aims to identify factors associated with an increased risk of 30-day readmission following hand surgery in the pediatric population. Our hypothesis is that pediatric hand procedures are safe procedures with low readmission rates.

The ACS NSQIP-P databases are a case-mix–adjusted repository of pediatric surgical cases from a variety of subspecialties that contain aggregate data on 30-day complications following surgical procedures for pediatric patients (ages 0-18 years). Data are abstracted by trained surgical clinical reviewers from more than 50 participating children’s hospitals, and outcomes are assessed for 30 days following the procedure. Data acquisition and maintenance have been described in detail elsewhere.¹⁵ The ACS NSQIP-P is a de-identified database compliant with the Health Insurance Portability and Accountability Act (HIPAA) and is therefore exempt from institutional review board review.

The 2012-2017 NSQIP-P databases were queried for pediatric patients who underwent procedures with hand-specific Current Procedural Terminology (CPT) codes. CPT codes 26010-26989 are associated with procedures on the hand and fingers and were queried in the NSQIP-P databases. Within this range referring to hand-specific procedures, 6 codes were identified in the databases. Patient demographics, operative details, readmission, and complication data were extracted. Variables assessed included sex, race, age, weight, gestational age at birth, American Society of Anesthesiologists (ASA) classification, inpatient or outpatient surgery, emergent or elective surgery, specialty performing the surgery, wound classification, anesthesia and operative times, any unplanned readmission, time to readmission, reason for readmission, reoperation, the presence of any complication, and wound complication. Incomplete charts (those not containing weight or ASA classification) were excluded. ASA classification was used as a surrogate for medical comorbidities.

The primary outcome assessed was unplanned readmission. Secondary outcomes included reasons for readmission and wound complications. Univariate analyses were performed using Fisher’s exact test, Wilcoxon rank-sum test, Chi-square, and analysis of variance as appropriate to assess the incidence and risk factors for readmission. Variables significant at a level of P < .20 were used to create a bidirectional, stepwise logistic regression model. Secondary outcomes were assessed with univariate analyses only. Statistical analysis was performed using STATA version 14 (College Station, TX). Statistical significance was defined as P < .05.

In the 2012-2017 databases, there were 6688 patients who underwent hand-specific procedures. Eighty-eight patients were excluded for incomplete data, leaving 6600 patients available for analysis. The characteristics of the cohort are shown in Table 1. Of the 181 procedural codes that were reported by the NSQIP-P, 6 hand-specific CPT codes were identified as the primary billing procedure (Table 2), including “tendon sheath incision” (26055), “syndactyly repair with flaps” (26560), “syndactyly repair with flaps and grafts” (26561), “complex syndactyly repair” (26562), “polydactyly repair” (26587), and “phalangeal fracture treatment” (26727). The most performed hand procedure was “tendon sheath incision” (26055), which accounted for 47.6% of procedures, followed by “polydactyly repair” (26587), which accounted for 26.3% of the procedures. Of the patients undergoing hand procedures, 64.7% received their care by Orthopedic Surgery, with Plastic Surgery and “Other” surgical specialty accounting for 33.3% and 2% of cases, respectively.

Of the study cohort, 45 were readmitted, giving an overall readmission rate of 0.68%. The median time to readmission was 12 days. On univariate analysis (Table 1), patient factors that were associated with readmission included younger age (31.9 vs 38.4 months, P = .04), smaller size (30.3 vs 34.6 pounds, P = .01), prematurity (20.0% vs 7.6%, P < .01), and an ASA class of 2 or greater (P < .01). Significant operative factors (Table 2) included inpatient admission at index operation (20.0% vs 3.5%, P < .01) and longer anesthesia (158.3 vs 99.9 minutes, P < .01) and operative times (106.2 vs 57.5 minutes, P < .01). CPT code was identified on initial frequency testing to be significantly associated with readmission (P < .01); however, post-hoc analysis failed to demonstrate significant differences on pairwise comparisons.

On multivariate analysis (Table 3), patient level factors associated with readmission included ASA class 3 (OR 3.95, 95% CI 1.57-9.34) or class 4 (OR 25.80, 95% CI 3.47-118.14) and prematurity (OR 2.18, 95% CI 0.96-4.51). Inpatient surgery had an odds ratio of being readmitted of 2.95 (95% CI 1.19-6.68). Increased operative time had a statistically significant but clinically unimportant effect on the odds of being readmitted (OR 1.00, 95% CI 1.00-1.01).

The NSQIP-P databases capture the reported reasons for readmission (Table 4). “Unplanned readmission” accounted for 40 of the 45 (88.9%) readmissions . Among patients who were readmitted, 16 (35.6%) were readmitted for a reason related to the prior surgery, including 8 related to a wound complication and 1 for significant bleeding. Increased anesthesia time was a significant predictor of reoperation (100.2 vs 142.8 minutes, P = .03).

Significant predictors of wound complications are reported in Table 5. One hundred twenty-four (1.9%) patients experienced a wound complication. The procedure most associated with wound complications was “complex syndactyly repair” (26562, 4.7%), and the least associated procedure was “phalangeal fracture treatment” (26727, 0.7%). The significant patient factors that predicted wound complications included younger age (28.9 vs 38.5 months, P < .01) and lower weight (29.4 vs 34.6 pounds, P < .01). Increased anesthesia and operative times were also associated with readmission (124.6 vs 99.8 min, P < .01; 80.8 vs 57.4 min, P < .01, respectively). Six (4.8%) patients with a wound complication required reoperation.

This study leverages 6 years of multi-institutional, aggregated data of 6600 patients from the NSQIP-P to quantify the risk of readmission for pediatric hand surgery. Overall, we found a 0.68% readmission rate. Pediatric hand surgery, for both congenital and traumatic indications, is safe and associated with a very low risk of 30-day readmission and a low complication profile.

The ACS NSQIP-P was piloted in 2008 and now collects 94 data points for 30-day outcomes for most pediatric surgical subspecialties.¹⁵ Several publications have used the adult or pediatric NSQIP database to assess 30-day outcomes within hand surgery, which are summarized in Table 6. Three prior studies have assessed the readmission and complication rates of pediatric patients undergoing upper extremity surgery. Thibaudeau and colleagues¹⁶ queried the 2012-2014 NSQIP-P databases for all upper extremity procedures and found a 0.78% readmission rate. McQuillan and colleagues¹⁷ examined the same databases for procedures related to congenital hand differences and found a readmission rate of 0.30%. That group furthermore described an increased risk of complications among patients undergoing “Repair of syndactyly each web space; complex, involving bone, nails, etc” (CPT code 25452).¹⁷ This outcome mirrors the finding in our study, with procedures associated with this CPT code result in higher rates of readmission. Chouairi et al¹⁸ assessed CPT codes related to syndactyly reconstruction specifically, using the 2012-2016 databases. In addition to comparing complication rates between simple and complex syndactyly, they published readmission rates of 0.8% and 2.1%, respectively.¹⁸ The overwhelming majority of complications in this study were related to surgical site complications. Our study adds to these previous studies the combined experience of traumatic and congenital hand surgeries over 6 years. Overall, the reported rates of readmission are low, and pediatric hand surgery remains very safe. However, future quality improvement efforts should be aimed at decreased surgical site complications.

The findings of the present study compare favorably with similar studies conducted using the adult NSQIP databases.^2-7 Two studies have looked at hand-specific CPT codes in adults using the 2006-2011 and 2011-2014 NSQIP databases.^2,4 Between 2006 and 2011, the adult NSQIP databases included 10 646 patients; however, the overall readmission rate was not reported. Subsequently, the 2011-2014 databases included 23 613 patients who underwent outpatient hand surgery, with a 0.88% readmission rate.² The 2011-2015 study found that patient medical factors contributed significantly to the risk of readmission, similar to findings in the pediatric population. Noureldin et al⁵ conducted a similar study looking at unplanned readmissions for hand- and elbow-specific CPT codes and found a 1.2% unplanned readmission rate. The readmitted patients were more likely higher ASA class and a lower preoperative functional status.⁵ Several studies have assessed the safety of specific procedures within upper extremity surgery. Studies on peripheral nerve surgery and corrective osteotomies of the forearm demonstrate readmission rates around 2%.^3,7 Recently, Shah and colleagues⁶ assessed 11 years’ worth of carpometacarpal arthroplasty cases and demonstrated a far lower readmission rate of 0.27%.

The present study identified several specific factors that were associated with readmission. The collective NSQIP-P data suggests an increased complication profile associated with complex syndactyly repair. The hallmark of complex syndactyly is abnormal skeletal or cartilaginous attachments, most commonly manifested as side-to-side fusion at the distal phalangeal tuft. These abnormalities also correlate with anomalies of the neurovascular bundles, placing these structures at risk during reconstruction.¹⁹ In addition, syndromic patients who can have higher anesthetic and perioperative risks may fall under this CPT code. Additionally, surgeries associated with an inpatient stay were 3 times more likely to be readmitted. McQuillan et al’s study¹⁷ focused specifically on congenital hand differences and did not find an association between admission status and subsequent readmission. This difference in these findings may be accounted for by the addition of patients undergoing tendon sheath incision and fracture fixation, which together constituted 49.9% of the present cohort. With respect to trauma specifically, some literature suggests that pediatric hand trauma in general is associated with a higher complication rate, but the collective administrative data in the NSQIP-P databases suggests that treatment of pediatric hand trauma is very safe.²⁰ In addition, case urgency was not associated with readmission. Finally, the findings of this multivariate model support the idea that the overall health of a child is a better predictor of readmission than the surgery itself.

The NSQIP-P databases solicits comments regarding the association between a procedure and subsequent readmission, although this information is variably available. In the present cohort, the readmission was considered to be related to the prior procedure about half of the time, with the other half representing medical problems unrelated to surgery. This finding suggests that readmissions due to technical or wound complications are exceptionally rare in pediatric hand surgery. Among the stated reasons for readmission, surgical site infection (SSI) was the most frequently cited. The present study revealed wound complications as the main surgical cause for readmission in the pediatric hand population, consistent with other studies in both pediatric and adult upper extremity surgery.² Specifically, variables suggestive of more significant medical problems, such as readmission, were all factors contributing to an increased risk of SSI.

This study was limited by several factors. Many of the readmission statistics did not include the reason for readmission, which would have offered more insight into risk factors of readmission. Also, the low readmission rate resulted in a small sample size from which to glean information regarding complications, limiting the ability to correlate readmission and complications with potential causes. In addition, the NSQIP-P database has been criticized for inaccurate data extraction when compared with departmentally collected data.²¹ Finally, retrospective, administrative data have inherent limitations when applied to clinical outcomes, as they lack granularity to assess patients at a detailed level. Future studies may include prospective examination of the impact of tourniquet use and wound care education interventions in the postoperative period.

Overall, pediatric hand surgery is associated with a very low rate of 30-day readmission. Higher ASA class and inpatient surgery increase this rate for patients. Complex syndactyly repair is associated with a higher risk of readmission than other hand procedures. This information is useful for surgical planning, postoperative care and appointments, preoperative optimization of expectations, counseling of parents, and nutritional optimization of the pediatric patient.

Affiliations: ¹Division of Plastic and Reconstructive Surgery, Department of Surgery, University of Louisville, Louisville, KY; ²University of Louisville School of Medicine, Louisville, KY; ³Division of General Surgery, Department of Surgery, Rutgers Robert Wood Johnson Medical School, New Brunswick, NJ; ⁴Division of Plastic and Reconstructive Surgery, Department of Surgery, Massachusetts General Hospital and Harvard Medical School, Boston, MA

Correspondence: Milind Kachare, MD; milind.kachare@louisville.edu

Ethics: Written informed consent was obtained from the patients for publication of this manuscript and accompanying images.

Disclosures: This study received no means of outside funding. The authors report no known or perceived conflicts of interest regarding the material presented in this manuscript.