The Relationship Between Relative Value Units and Outcomes: A Multivariate Analysis of Plastic Surgery Procedures

Patient-related risk factors for complications

Multivariate regression modeling identified a number of patient-associated variables that were significantly associated with an increased risk of complications or reoperation. Many of these factors are not surprising and in line with previously published reports, including increasing age, ASA classification, and BMI, and also chronic steroid use. Increasing age, BMI, and ASA classification are well-known risk factors for complications in various surgical specialties, including plastic and reconstructive surgery.^21-24 Chronic steroid use is thought to impair wound healing and increase the risk of postoperative wound infection.^25-27 Other factors associated with increased risks of complications or reoperation included poor baseline functional status, preoperative wound infections or sepsis, metastatic malignancy, history of transient ischemic attack (TIA), poor baseline functional status, previous coronary revascularization, and recent surgery. Although these variables have not been extensively studied as independent risk factors for postoperative complications,^14,28,29 they suggest severe underlying pathologies, such as systemic illness, vascular insufficiency or advanced atherosclerosis, immunosuppression, malnutrition, autonomic instability, or other pathologic processes that may predict poorer surgical outcomes in these patients.

Interestingly, a history of previous cardiac surgery was found to predict a lower rate of reoperation. One potential explanation for this might be reluctance of surgeons to reoperate on a patient with a significant history of heart disease in a situation where a healthier patient might benefit from reoperation. For example, a wound dehiscence that might be closed by primary or delayed primary intent in a healthy patient might be left to heal secondarily in a patient with significant cardiac comorbidities to avoid the risk of additional anesthetic.

Intraoperative risk factors for complications

As might be predicted, ambulatory surgical procedures were associated with a dramatic 66.6% reduction in the odds of overall complications and a 52.3% reduction in the odds of surgical site complications compared with inpatient procedures. This indirectly supports the idea that more complex surgeries, which are usually undertaken in an inpatient setting, may lead to poorer outcomes. However, factors such as increased risk of nosocomial infections and higher prevalence of multidrug resistant organisms in inpatient settings may confound this analysis.^29,30

Resident participation in surgery predicted a 35.3% increase in the odds of overall complications and 40.3% increase in the odds of surgical site complications. This unexpected finding has several potential explanations. One logical explanation might be the prolongation of operative times associated with intraoperative house-staff teaching, for both surgical and anesthesiology residents. Another might be inexperience and technical error on the part of the trainee. This is certainly an area for future investigation across all surgical disciplines.

Predictors of mortality

As noted earlier, many of the factors associated with an increased risk of mortality are those associated with poorer baseline health and functional status. The relatively small number of mortalities in this series (n = 50 [0.33%]) precluded a robust statistical analysis of positive and negative predictors for this outcome.

Significance of RVUs as a predictor of complications

While previous studies have used RVUs as a surrogate for surgical complexity in individual procedures, it is unclear whether summation of RVUs from concurrent procedures can similarly substitute for overall surgical complexity. However, for patients with multiple procedures, the primary procedure, in this case, all plastic surgical procedures, contributed 55% of total RVUs included in this analysis. In addition, subanalysis of only patients undergoing multiple procedures indicated that surgical complexity, as measured by total RVUs, persisted as significant risk factor for complications (data not shown). Thus, it is likely that total RVUs is a valid measure of surgical complexity for patients undergoing multiple procedures; however, this cannot be definitively stated from the present results.

It stands to reason that added surgical complexity will only increase the potential for postoperative complications; however, this conclusion is not rooted in strong evidence. Previous studies have demonstrated increased complications when complexity has been added to traditional procedures but have not quantitatively described the magnitude of these changes. Thus, the existing literature offers only limited insight into the relationship between surgical complexity and outcomes in plastic and reconstructive surgery. For example, lipoabdominoplasty shows a higher rate of seroma formation than traditional abdominoplasty³¹; the extent of tissue excision directly corresponds to seroma rates in body contouring procedures³²; and facelifts performed in conjunction with additional procedures have demonstrated an increased risk of deep venous thrombosis.³³ Similarly, immediate breast reconstruction, which is performed in conjunction with mastectomy, may lead to higher complications rates than delayed reconstructions performed as a separate procedure.^34-37 In addition to their limited scope, many of these existing comparisons are drawn from single-surgeon or single-institution series, incorporating a natural selection bias and reducing generalizability. The results of the current study are derived from numerous military, community, and academic institutions across the country, which decreases selection bias as a potential confounder and broadens the applicability of our results.

We found that a single unit increase in total RVUs increased the incidence of overall complications by 1.7% and the incidence of surgical site complications by 1.0% in the 30-day postoperative period. Increasing total RVUs, a surrogate for increasing surgical complexity, may lead to longer operative times and, therefore more prolonged exposure to anesthesia, with resultant increased risk of deep venous thrombosis, arrhythmia, and other related complications.^22,33,38,39 Linear regression indicated that a single unit increase in RVUs was associated with a 0.410-minute increase in total operative time. However, variability in total RVUs only explained 15.6% of operative time variability in this study, suggesting that this association does not tell the complete story, and other factors are likely contributing to these observed trends. Given that concomitant surgical procedures can often be performed in parallel, particularly when surgical house-staff are present to assist, it is not surprising that operative times do not increase linearly with increasing RVUs.

Although RVUs were a significant risk factor for overall and surgical site complications, they did not predict reoperations or increased mortality, suggesting that, while complex surgery may result in a higher overall rate of complications, a significant proportion of these complications might be successfully managed nonoperatively and do not significantly contribute to overall mortality.

Limitations and additional considerations

Limitations of this study include the relatively short follow-up data provided by the NSQIP database. History of radiotherapy to the involved anatomy was not tracked, despite being a well-known risk factor of various complications following reconstructive and cosmetic procedures. Although many complications, particularly those requiring reoperation, will tend to occur in the early postoperative period, several complications, including deep venous thrombosis/pulmonary embolism (DVT/PE), seroma and wound breakdown among others, may not clinically manifest within 30 days of surgery. Moreover, this initial study is limited to procedures performed primarily by plastic surgeons. In future iterations, we can also apply this methodology to multidisciplinary operative cases. Lastly, some procedures falling under the purview of plastic surgery are not covered by Medicare and, as such, are not assigned an RVU value (ie, RVU = 0), therefore the relative contribution of these procedures cannot be included in this analysis.

The authors recognize the potential implications of linking RVUs to complications in the era of outcomes-based reimbursement. It is important to note that the valuation of procedures via RVUs includes factors beyond procedural complexity, such as training time for a surgeon to reach technical proficiency and the intensity of work required to complete the procedure. Furthermore, complex procedures may incur an acceptable increase in the risk of complications to save time, provide greater patient benefits, and reduce costs.⁴⁰ Finally, it is plausible that high RVUs actually account for the costs associated with managing complications and that these costs vary by procedure. Thus, it would be too simplistic to argue an association with complications threatens the validity of current RVU assignments.

However, it is worthwhile to note that RVU productivity correlates with surgical department margins.⁴¹ Moreover, RVU productivity per unit time worked can vary dramatically across procedures and surgical specialty.⁴² The finding of RVUs as an independent risk factor for complications in our analysis raises an important health policy question: How should surgical complexity be considered in the reimbursement scheme? The answer to this question hangs in the balance of many current political agendas and will likely not be decided for some time. In the interim, further investigation of the impact of surgical complexity on postoperative complication rates in all surgical disciplines will help improve quality and safety of surgical treatment.