LACE+ Index as a Predictor of 90-Day Plastic Surgery Outcomes

April 9, 2020
Eric Winter, BS

,
Gregory Glauser, BS

,
Ian F. Caplan, BS

,
Stephen Goodrich, BS

,
Scott D. McClintock, PhD

,
Stephen J. Kovach III, MD

,
Joshua Fosnot, MD

,
Joseph M. Serletti, MD

,
Neil R. Malhotra, MD

Volume 26, Issue 04

Patients with higher LACE+ index scores have significantly greater risk of unplanned readmission, emergency department visits, and reoperation after plastic surgery.

ABSTRACT

Objectives: This study used coarsened exact matching to assess the ability of the LACE+ index to predict adverse outcomes after plastic surgery.

Study Design: Two-year retrospective study (2016-2018).

Methods: LACE+ scores were retrospectively calculated for all patients undergoing plastic surgery at a multicenter health system (N = 5744). Coarsened exact matching was performed to sort patient data before analysis. Outcomes including unplanned hospital readmission, emergency department visits, and reoperation were compared for patients in different LACE+ score quartiles (Q1, Q2, Q3, Q4).

Results: A total of 2970 patient procedures were matched during coarsened exact matching. Increased LACE+ score significantly predicted readmission within 90 days of discharge for Q4 versus Q1 (6.28% vs 1.91%; P = .003), Q4 versus Q2 (12.30% vs 5.56%; P <.001), and Q4 versus Q3 (13.84% vs 7.33%; P <.001). Increased LACE+ score also significantly predicted emergency department visits within 90 days for Q4 versus Q1 (9.29% vs 3.01%; P <.001), Q4 versus Q2 (11.31% vs 3.57%; P <.001), and Q4 versus Q3 (13.70% vs 8.48%; P = .003). Higher LACE+ score also significantly predicted secondary reoperation within 90 days for Q4 versus Q1 (3.83% vs 1.37%; P = .035), Q4 versus Q2 (5.95% vs 3.37%; P = .042), and Q4 versus Q3 (7.50% vs 3.26%; P <.001).

Conclusions: The results of this study demonstrate that the LACE+ index may be suitable as a prediction model for patient outcomes in a plastic surgery population.

Am J Manag Care. 2020;26(4):e113-e120. https://doi.org/10.37765/ajmc.2020.42838

Takeaway Points

The present study assessed the ability of the LACE+ index to identify patients at risk for adverse outcomes after plastic surgery:

  • Patients with the highest LACE+ scores were readmitted nearly 4 times more than patients with the lowest LACE+ scores. Similar results were observed for rates of emergency department visits and reoperation.
  • Coarsened exact matching was employed to attenuate the influence of confounding variables on analysis.
  • The LACE+ index may be used to identify patients at high risk who would benefit from targeted interventions designed to improve postdischarge outcomes.

Throughout the last several decades, healthcare costs in the United States have exploded.1 In 2018, healthcare accounted for $3.5 trillion in total spending and comprised nearly 18% of the overall gross domestic product.2 This increasingly unsustainable trend has become the subject of intense scrutiny, and many stakeholders have initiated programs and developed incentives in an attempt to rein in costs.3-5 In this context, unplanned hospital readmission—which accounts for more than $15 billion in spending annually—has become a particularly important metric.6 By proactively identifying and supporting high-risk patients, physicians may be able to increase care quality, reduce preventable readmissions, and eliminate unnecessary costs.7-9

Plastic surgery is an important area of focus due to increasing volume and complexity of procedures.10 Previous work has found that approximately 11% of patients undergoing plastic surgery will require readmission to the hospital; leading causes of readmission include surgical site infection, wound disruption, and bleeding.6,7,11 Certain populations have particularly high rates of readmission, including patients with a history of obesity, coronary intervention, and chronic obstructive pulmonary disease.6,7 Other factors, such as old age, recent surgery, and increased length of stay, have also been correlated with increased rates of hospital readmission.7,12 Interventions targeted at patients with these risk factors may be able to significantly improve postdischarge outcomes, but unfortunately there is currently no validated tool that is able to integrate this information to stratify patient risk of readmission.

The LACE (Length of stay, Acuity of admission, Charlson Comorbidity Index [CCI] score, and Emergency department [ED] visits in the past 6 months) index was developed to help provide a simple metric to predict 30-day readmission after hospital discharge.13 Prior work has tested the LACE index in several surgical and disease-specific medical populations, with varied results.14-19 The original creators of the LACE index subsequently modified it to create the LACE+ index, which incorporates other variables, including age, gender, and previous hospital admission.20

To date, there are no published data supporting the validity of the LACE+ index in a plastic surgery population. Additionally, the LACE+ index has never been used to predict patient outcomes more than 30 days after discharge. The present investigation employs coarsened exact matching to sort patient data.21 Using matched surgical experiences, this study assesses the ability of the LACE+ index to predict 90-day patient outcomes in a mixed-procedure plastic surgery population.

METHODS

Sample Selection

In this University of Pennsylvania institutional review board (IRB)—approved study, patients undergoing plastic surgical intervention at a multihospital, 1659-bed university health system were enrolled retrospectively over 2 years (January 1, 2016, to January 1, 2018). A waiver of informed consent was granted by the IRB because this study was considered to be of minimal risk to patients. There were 5744 plastic surgery procedures studied. Key data were acquired using the EpiLog tool—a nonproprietary data acquisition system created by the senior author of this article (N.R.M.). It was built and layered on top of the existing electronic health record (EHR) architecture to facilitate charting, workflow, quality improvement, and cost-reduction initiatives.22

Data Collection

Relevant patient information, including sex, admission type, length of stay, CCI score, recent ED visits, surgical history, and history of hospital admissions, was scored and incorporated into a composite LACE+ index value (—1 to 90) (Figure 1 [A]). After calculating LACE+ scores for all patients (eAppendix Figure [A] [eAppendix available at ajmc.com]), quartiles (Q1, Q2, Q3, Q4) were determined (eAppendix Figure [B]).

Patient characteristics not used for LACE+ score calculation were also documented, and they served as matching criteria (Figure 1 [B]). These included operative time, surgical cost, household income, insurance type, race, body mass index (BMI), patient class (inpatient, outpatient), and wound class (clean, contaminated, dirty). Data were binned (“coarsened”) to create discrete categories within each matching criterion.

Occurrence of readmission, ED visit, and reoperation within 90 days were recorded. Patient Current Procedural Terminology (CPT) billing codes were also documented to describe the composition of case types in the cohort, although this information was not used for matching or data analysis (eAppendix Table).

Matching

Coarsened exact matching was performed (Figure 2 [A]). Matches were sought between patients in different LACE+ score quartiles but with otherwise identical characteristics (matching criteria). During this process, patients with the highest LACE+ scores (Q4) were matched to patients with lower LACE+ scores (Q3, Q2, Q1). Unmatched patients from each LACE+ quartile were removed from the data set and not included in further analysis (Figure 2 [B]). The result was a series of matched groups (Q4-Q1, Q4-Q2, Q4-Q3). Groups in these pairs differed only with respect to LACE+ score and had otherwise identical patient composition.

Finally, outcomes for patients in each quartile (eg, Q4, Q1) were analyzed and compared for all matched groups (eg, Q4-Q1).

Statistical Analysis

Comparisons were made for Q4 versus Q1, Q4 versus Q2, and Q4 versus Q3. McNemar’s test was used to assess the ability of the LACE+ index to accurately predict outcome measures by comparing the means between quartiles. Univariate analysis was performed with significance set at P <.05.

RESULTS

Patient Characteristics

The prematch sample included all consecutive patients undergoing plastic surgery over 2 calendar years (N = 5744) (Table 1 [part A and part B]). Median BMI was 26.2. Median length of stay was 7.00 hours, and median operative time was 66 minutes for all patients. Median total cost of surgical supplies and implants was $278. Median household income for patients was $69,874 for the overall cohort.

Patient Outcomes

We matched 366 patients between Q4 and Q1 (n = 732; a 28.2% match rate), 504 patients between Q4 and Q2 (n = 1008; a 36.7% match rate), and 615 patients between Q4 and Q3 (n = 1230; a 44.8% match rate) (Figure 3; Table 2).

Prediction of Readmissions

Higher LACE+ scores were found to significantly predict risk of readmission within 90 days of discharge for Q4 versus Q1 (6.28% vs 1.91%; odds ratio [OR], 3.67; 95% CI, 1.49-9.04; P = .003), Q4 versus Q2 (12.30% vs 5.56%; OR, 2.42; 95% CI, 1.50-3.89; P <.001), and Q4 versus Q3 (13.84% vs 7.33%; OR, 2.29; 95% CI, 1.50-3.49; P <.001).

Increased LACE+ scores were also found to significantly predict risk of readmission in the 30-to-90-day postdischarge window for Q4 versus Q1 (4.37% vs 1.09%; OR, 5.00; 95% CI, 1.41-26.9; P = .008), Q4 versus Q2 (5.56% vs 2.58%; OR, 2.25; 95% CI, 1.14-4.44; P = .016), and Q4 versus Q3 (7.00% vs 3.58%; OR, 2.11; 95% CI, 1.22-3.64; P = .006).

Prediction of ED Visits

Higher LACE+ scores were found to significantly predict risk of ED visits within 90 days of discharge for Q4 versus Q1 (9.29% vs 3.01%; OR, 3.30; 95% CI, 1.63-6.69; P <.001), Q4 versus Q2 (11.31% vs 3.57%; OR, 3.44; 95% CI, 1.97-5.99; P <.001), and Q4 versus Q3 (13.70% vs 8.48%; OR, 1.76; 95% CI, 1.21-2.57; P = .003).

Increased LACE+ scores were also found to significantly predict risk of ED visits in the 30-to-90-day postdischarge window for Q4 versus Q1 (4.10% vs 1.09%; OR, 3.75; 95% CI, 1.19-15.5; P = .019), Q4 versus Q2 (5.16% vs 1.39%; OR, 4.17; 95% CI, 1.71-10.2; P <.001), and Q4 versus Q3 (7.67% vs 2.77%; OR, 3.31; 95% CI, 1.78-6.15; P <.001).

Prediction of Reoperation

Higher LACE+ scores were found to significantly predict risk of reoperation within 90 days of discharge for Q4 versus Q1 (3.83% vs 1.37%; OR, 4.00; 95% CI, 1.08-22.1; P = .035), Q4 versus Q2 (5.95% vs 3.37%; OR, 1.93; 95% CI, 1.01-3.68; P = .042), and Q4 vs Q3 (7.50% vs 3.26%; OR, 2.63; 95% CI, 1.48-4.67; P <.001).

Increased LACE+ scores were not found to significantly predict risk of reoperation in the 30-to-90-day postdischarge window for Q4 versus Q1 (2.19% vs 0.55%; OR, 4.00; 95% CI, 0.79-38.7; P = .109), Q4 versus Q2 (3.57% vs 1.98%; OR, 1.89; 95% CI, 0.84-4.23; P = .117), or Q4 versus Q3 (4.08% vs 2.28%; OR, 1.85; 95% CI, 0.94-3.63; P = .071).

DISCUSSION

The present study sought to determine if the LACE+ index—a previously developed tool validated in other specialties—could be used to predict patient outcomes in a mixed-procedure plastic surgery population.

After comparing data for all quartiles, results suggest that LACE+ score is able to reliably predict outcomes within 90 days of discharge. Patients with a Q4 LACE+ score were 2.3 to 3.7 times more likely to be readmitted than patients with lower LACE+ scores. Patients with Q4 scores also had 1.8 to 3.4 times the risk of ED visits and 1.9 to 4.0 times the risk of reoperation compared with patients in lower LACE+ score quartiles. LACE+ also predicted outcomes in the 30-to-90-day postdischarge window. Patients with Q4 LACE+ scores were found to have 2.1 to 5.0 times the risk of readmission and 3.3 to 4.2 times the risk of ED visits compared with patients in lower LACE+ score quartiles.

The LACE+ index was originally developed to predict patient outcomes within 30 days of hospital discharge. The results of the present work suggest that the LACE+ index is, in fact, able to reliably predict patient outcomes within 90 days of discharge. Notably, data analysis revealed that increased LACE+ score is significantly predictive of patient outcomes in the 30-to-90-day postdischarge window. This demonstrates that the predictive power of the LACE+ index in this 90-day time frame is not achieved by simply predicting outcomes within the first 30 days after discharge.

This expanded predictive window may facilitate clinician and hospital-led efforts to more effectively manage the health of patients undergoing plastic surgery in the months after discharge. As value-based contracts and risk-sharing agreements become more commonplace, this sort of long-term population health management will become increasingly important.

Management of Adverse Outcomes&#8203;&#8203;&#8203;&#8203;&#8203;&#8203;&#8203;

Although the present study did not seek to investigate the precise causes of patient readmission, previous work has identified surgical site infection, among other wound-related complications, as a leading cause of readmission after plastic surgery.6,7,11 Although data are limited on isolated plastic surgery populations, a study of patients undergoing vascular surgery by Wiseman et al found that nonmodifiable factors, such as chronic disease and emergency case status, were significant predictors of potentially preventable, readmission-causing complications (eg, surgical site infection).8 Importantly, these same factors inform elements of the LACE+ index.21

Many interventions have been proposed and tested to reduce readmission after hospital discharge. Dawes et al identified infection and dehydration as significant drivers of preventable hospital readmission in a general surgery population.23 These authors suggested that improved outpatient management and closer postdischarge follow-up might be able to prevent these problems and decrease hospital readmissions.23

The existing literature strongly supports this approach. A number of studies have demonstrated that improved discharge planning and increased postdischarge follow-up is associated with lower risk of surgical site infection and lower rates of readmission, particularly for high-risk patients.9,24-26 A randomized study by Naylor and McCauley focused on patients hospitalized with surgical cardiac conditions. They found that a combination of robust discharge planning, proactive telephone outreach, question hotlines, and home nurse visits was able to significantly reduce readmissions.27 Bisognano and Boutwell’s review of interventions for patients with heart failure supported these interventions and suggested that a similar suite of programs was able to significantly reduce readmissions in that population.28 The value of this multipronged approach has been reaffirmed by subsequent research, including a systematic review of 43 readmission-reduction programs by Hansen et al that suggested that some interventions, such as telephone outreach, are particularly effective when targeted at high-risk populations.29

LACE+ in Clinical Practice

The results of the present work suggest that the LACE+ index is able to risk-stratify an entire plastic surgery population, independent of any specific case/procedure type. This observation has the potential to significantly reduce the administrative burden associated with integrating the LACE+ index into clinical practice. LACE+ scores can be automatically calculated for each patient in the EHR, and supportive resources can be instinctively offered to any patient with a high LACE+ score, regardless of the type of operation they received. This streamlines the clinical workflow and prevents the need for staff to manually classify procedure types and choose which patients receive an intervention.

The LACE+ algorithm may be incredibly valuable in clinical environments with limited resources to support patients and prevent adverse outcomes after hospital discharge. Our results suggest that the LACE+ index may allow clinicians to identify high-risk patients who would benefit from supportive resources, such as home nursing, outreach calls, and follow-up office visits. Put differently, the LACE+ index may help physicians avoid unnecessary and cost-ineffective interventions for low-risk patients who are not expected to experience adverse outcomes. Without the LACE+ index, this data-driven resource allocation strategy may not be possible.

In the near future, risk-prediction algorithms will likely become an increasingly important component of effective population health management. To this end, the LACE and LACE+ indices helpfully integrate multiple discrete variables into a single number, although their actual predictive value has fallen short in several prior studies.15,16,30

The results of the present study suggest that the LACE+ index may be a valuable and accurate indicator of risk for readmission and ED visits following plastic surgery. The use of coarsened exact matching in the current work provides additional assurance that this predictive power is not a result of confounding variables. The present study also indicates that the LACE+ index may also predict risk of reoperation, although this should be subjected to further study.

Limitations

This study relied on accurate recording of patient readmission information into the EHR. As a result, it is possible that the rate of readmission in this work was underreported. Although this is a potential confounding variable, all enrolled patients did visit the operative surgeon during the surgical follow-up window. During this encounter, the EpiLog tool captured data on all clinical interactions in the postoperative window, regardless of location or hospital system. Other errors in data recording may have been present and influenced analysis. However, the coarsened exact matching process should randomly distribute these hypothetical inaccuracies among study groups. Finally, the generalizability of this study may be limited because the sample population was drawn from a single institution.

This study used the statistical method of coarsened exact matching. Patients were matched on many key criteria, but it is possible that additional unknown/unstudied characteristics would have facilitated more perfect matching between quartiles. The matching criteria employed in the present study were used because of prior work correlating these factors with patient outcomes. Studies by Connolly et al and LaPar et al demonstrated that patients who are low-income, uninsured, and insured by Medicaid have higher rates of postsurgical inpatient mortality.31,32 A number of other studies have reported that BMI (particularly obesity), history of previous operation, minority race, and case type are all independently associated with increased postoperative complications.33-40 Work by Chu et al and Kim et al also showed that increased length of surgery is associated with higher risk of morbidity and longer stays in the intensive care unit.41,42

Although CPT codes were documented, this information was not used to match patients based on specific plastic surgery procedures. In this article, 512 unique CPT codes (in 108 CPT code ranges) were used to bill for services. Unfortunately, the sample population of 5744 patients was simply not large enough to support matching with so many discrete subgroups. If CPT codes were used as matching criteria, it is likely that only a handful of matched patients would have been identified. Because of this limitation, related factors, including operating time, surgical cost, patient class (inpatient, outpatient), and wound class (clean, clean contaminated, dirty), were used as matching criteria. These variables were chosen to control for procedure type without severely bottlenecking matching; it is possible that these matching criteria did not effectively distribute procedure types among groups.

CPT codes were also not used to perform subgroup analysis of specific plastic surgery procedure types. The intent of this work was to assess the ability of LACE+ to risk-stratify an entire plastic surgery population, as may be done at a departmental or institutional level. This study did not seek to investigate LACE+ risk prediction for specific types of plastic surgeries; this is a potential area for follow-up research.

Future studies should test the LACE+ index in a larger, perhaps multicenter, plastic surgery population. This may increase statistical power and allow for validation of the LACE+ index in specific procedure types. Work should also be undertaken to strengthen the correlation between LACE+ score and specific causes of patient readmission. Finally, prospective work should be initiated to assess the ability of the LACE+ index to inform resource engagement and improve quality measures.

CONCLUSIONS

This study represents one of the first investigations assessing the validity of the LACE+ index in a plastic surgery population. It is also the first to use the LACE+ index to predict patient outcomes within 90 days of discharge. The results of this study suggest that this tool may be able to accurately assess risk of hospital readmission, ED utilization, and reoperation after plastic surgical intervention. Coarsened exact matching strengthens this assessment by controlling for potential confounding variables. Future research should prospectively test the ability of the LACE+ index to effectively target resources to support high-risk patients undergoing plastic surgery.

Acknowledgments

The authors would like to acknowledge the support of The Neurosurgery Quality Improvement Initiative (NQII) EpiLog Project and the Bernadette and Kevin McKenna Family Research Fund in support of this project.Author Affiliations: Department of Neurosurgery (EW, GG, IFC, NRM) and Division of Plastic Surgery, Department of Surgery (SJK, JF, JMS), Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA; McKenna EpiLog Fellowship in Population Health, University of Pennsylvania (SG, NRM), Philadelphia, PA; West Chester Statistical Institute and Department of Mathematics, West Chester University (SG, SDM), West Chester, PA.

Source of Funding: Bernadette and Kevin McKenna Family Research Fund.

Author Disclosures: The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (EW, NRM); acquisition of data (SG, SDM, SJK, JF, JMS, NRM); analysis and interpretation of data (GG, SG, SDM, NRM); drafting of the manuscript (EW, GG, IFC, NRM); critical revision of the manuscript for important intellectual content (EW, GG, IFC, SG, SDM, SJK, JF, JMS, NRM); statistical analysis (IFC, SG, SDM); provision of patients or study materials (SJK, JF, JMS, NRM); administrative, technical, or logistic support (GG, NRM); and supervision (NRM).

Address Correspondence to: Neil R. Malhotra, MD, Department of Neurosurgery, Hospital of the University of Pennsylvania, Silverstein Pavilion, 3400 Spruce St, 3rd Floor, Philadelphia, PA 19104. Email: NRM@uphs.upenn.edu.REFERENCES

1. National health expenditure data: historical. CMS website. cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/NationalHealthExpendData/NationalHealthAccountsHistorical.html. Updated December 17, 2019. Accessed December 19, 2019.

2. NHE fact sheet. CMS website. cms.gov/research-statistics-data-and-systems/statistics-trends-and-reports/nationalhealthexpenddata/nhe-fact-sheet.html. Updated December 5, 2019. Accessed December 19, 2019.

3. Begley C, Hall J, Shenoy A, et al. Design and implementation of the Texas Medicaid DSRIP program. Popul Health Manag. 2017;20(2):139-145. doi: 10.1089/pop.2015.0192.

4. Gates A, Rudowitz R, Guyer J. An overview of Delivery System Reform Incentive Payment (DSRIP) waivers. Kaiser Family Foundation website. kff.org/medicaid/issue-brief/an-overview-of-delivery-system-reform-incentive-payment-waivers. Published September 29, 2014. Accessed March 11, 2020.

5. Kocher RP, Adashi EY. Hospital readmissions and the Affordable Care Act: paying for coordinated quality care. JAMA. 2011;306(16):1794-1795. doi: 10.1001/jama.2011.1561.

6. Jain U, Salgado C, Mioton L, Rambachan A, Kim JYS. Predictors of readmission after inpatient plastic surgery. Arch Plast Surg. 2014;41(2):116-121. doi: 10.5999/aps.2014.41.2.116.

7. Fischer JP, Wes AM, Nelson JA, Serletti JM, Kovach SJ. Factors associated with readmission following plastic surgery: a review of 10,669 procedures from the 2011 American College of Surgeons National Surgical Quality Improvement Program data set. Plast Reconstr Surg. 2013;132(3):666-674. doi: 10.1097/PRS.0b013e31829acc8c.

8. Wiseman JT, Fernandes-Taylor S, Barnes ML, et al. Predictors of surgical site infection after hospital discharge in patients undergoing major vascular surgery. J Vasc Surg. 2015;62(4):1023-1031.e5. doi: 10.1016/j.jvs.2015.04.453.

9. Rochon M, Makhecha S, Morais C, et al. Quality improvement approach to reducing readmission for surgical site infection. Wounds UK. 2016;12(2):26-31.

10. 2017 plastic surgery statistics report. American Society of Plastic Surgeons website. plasticsurgery.org/documents/News/Statistics/2017/plastic-surgery-statistics-full-report-2017.pdf. Accessed March 11, 2020.

11. Khavanin N, Alghoul MS, Blough JT, Vu MM, Kim JYS. Underlying reasons and timing associated with readmission following plastic surgery: analysis of a national surgical database. J Plast Reconstr Aesthet Surg. 2016;69(11):1568-1571. doi: 10.1016/j.bjps.2016.08.021.

12. Tahiri Y, Fischer JP, Wink JD, et al. Analysis of risk factors associated with 30-day readmissions following pediatric plastic surgery: a review of 5376 procedures. Plast Reconstr Surg. 2015;135(2):521-529. doi: 10.1097/PRS.0000000000000889.

13. van Walraven C, Dhalla IA, Bell C, et al. Derivation and validation of an index to predict early death or unplanned readmission after discharge from hospital to the community. CMAJ. 2010;182(6):551-557. doi: 10.1503/cmaj.091117.

14. Fanari Z, Elliott D, Russo CA, Kolm P, Weintraub WS. Predicting readmission risk following coronary artery bypass surgery at the time of admission. Cardiovasc Revasc Med. 2017;18(2):95-99. doi: 10.1016/j.carrev.2016.10.012.

15. Cotter PE, Bhalla VK, Wallis SJ, Biram RWS. Predicting readmissions: poor performance of the LACE index in an older UK population. Age Ageing. 2012;41(6):784-789. doi: 10.1093/ageing/afs073.

16. Wang H, Robinson RD, Johnson C, et al. Using the LACE index to predict hospital readmissions in congestive heart failure patients. BMC Cardiovasc Disord. 2014;14:97. doi: 10.1186/1471-2261-14-97.

17. Hakim MA, Garden FL, Jennings MD, Dobler CC. Performance of the LACE index to predict 30-day hospital readmissions in patients with chronic obstructive pulmonary disease. Clin Epidemiol. 2018;10:51-59. doi: 10.2147/CLEP.S149574.

18. El Morr C, Ginsburg L, Nam VS, Woollard S, Hansen B. Analyzing readmissions patterns: assessment of the LACE tool impact. Stud Health Technol Inform. 2016;223:25-30. doi: 10.3233/978-1-61499-645-3-25.

19. Damery S, Combes G. Evaluating the predictive strength of the LACE index in identifying patients at high risk of hospital readmission following an inpatient episode: a retrospective cohort study. BMJ Open. 2017;7(7):e016921. doi: 10.1136/bmjopen-2017-016921.

20. van Walraven C, Wong J, Forster AJ. LACE+ index: extension of a validated index to predict early death or urgent readmission after hospital discharge using administrative data. Open Med. 2012;6(3):e80-e90.

21. Iacus SM, King G, Porro G. Causal inference without balance checking: coarsened exact matching. Polit Anal. 2012;20(1):1-24. doi: 10.1093/pan/mpr013.

22. Gawande A. Why doctors hate their computers. The New Yorker website. newyorker.com/magazine/2018/11/12/why-doctors-hate-their-computers. Published November 5, 2018. Accessed February 6, 2019.

23. Dawes AJ, Sacks GD, Russell MM, et al. Preventable readmissions to surgical services: lessons learned and targets for improvement. J Am Coll Surg. 2014;219(3):382-389. doi: 10.1016/j.jamcollsurg.2014.03.046.

24. Hernandez AF, Greiner MA, Fonarow GC, et al. Relationship between early physician follow-up and 30-day readmission among Medicare beneficiaries hospitalized for heart failure. JAMA. 2010;303(17):1716-1722. doi: 10.1001/jama.2010.533.

25. Baack Kukreja J, Kamat AM. Strategies to minimize readmission rates following major urologic surgery. Ther Adv Urol. 2017;9(5):111-119. doi: 10.1177/1756287217701699.

26. Burke RE, Coleman EA. Interventions to decrease hospital readmissions: keys for cost-effectiveness. JAMA Intern Med. 2013;173(8):695-698. doi: 10.1001/jamainternmed.2013.171.

27. Naylor MD, McCauley KM. The effects of a discharge planning and home follow-up intervention on elders hospitalized with common medical and surgical cardiac conditions. J Cardiovasc Nurs. 1999;14(1):44-54. doi: 10.1097/00005082-199910000-00006.

28. Bisognano M, Boutwell A. Improving transitions to reduce readmissions. Front Health Serv Manage. 2009;25(3):3-10.

29. Hansen LO, Young RS, Hinami K, Leung A, Williams MV. Interventions to reduce 30-day rehospitalization: a systematic review. Ann Intern Med. 2011;155(8):520-528. doi: 10.7326/0003-4819-155-8-201110180-00008.

30. Robinson R, Hudali T. The HOSPITAL score and LACE index as predictors of 30 day readmission in a retrospective study at a university-affiliated community hospital. PeerJ. 2017;5:e3137. doi: 10.7717/peerj.3137.

31. Connolly TM, White RS, Sastow DL, Gaber-Baylis LK, Turnbull ZA, Rong LQ. The disparities of coronary artery bypass grafting surgery outcomes by insurance status: a retrospective cohort study, 2007-2014. World J Surg. 2018;42(10):3240-3249. doi: 10.1007/s00268-018-4631-9.

32. LaPar DJ, Bhamidipati CM, Harris DA, et al. Gender, race, and socioeconomic status affects outcomes after lung cancer resections in the United States. Ann Thorac Surg. 2011;92(2):434-439. doi: 10.1016/j.athoracsur.2011.04.048.

33. Smits A, Lopes A, Das N, et al. Surgical morbidity and clinical outcomes in ovarian cancer — the role of obesity. BJOG. 2016;123(2):300-308. doi: 10.1111/1471-0528.13585.

34. Wilson JR, Tetreault LA, Schroeder G, et al. Impact of elevated body mass index and obesity on long-term surgical outcomes for patients with degenerative cervical myelopathy: analysis of a combined prospective dataset. Spine (Phila Pa 1976). 2017;42(3):195-201. doi: 10.1097/BRS.0000000000001859.

35. Causey MW, McVay D, Hatch Q, et al. The impact of race on outcomes following emergency surgery: an American College of Surgeons National Surgical Quality Improvement Program assessment. Am J Surg. 2013;206(2):172-179. doi: 10.1016/j.amjsurg.2012.11.022.

36. Stone JS, Fukuoka H, Weinreb RN, Afshari NA. Relationship between race, insurance coverage, and visual acuity at the time of cataract surgery. Eye Contact Lens. 2018;44(6):393-398. doi: 10.1097/ICL.0000000000000443.

37. Akbarian-Tefaghi H, Kalakoti P, Sun H, et al. Impact of hospital caseload and elective admission on outcomes after extracranial-intracranial bypass surgery. World Neurosurg. 2017;108:716-728. doi: 10.1016/j.wneu.2017.09.082.

38. Solsky I, Friedmann P, Muscarella P, In H. Poor outcomes of gastric cancer surgery after admission through the emergency department. Ann Surg Oncol. 2017;24(5):1180-1187. doi: 10.1245/s10434-016-5696-z.

39. Radcliff K, Jalai C, Vira S, et al. Two-year results of the Prospective Spine Treatment Outcomes Study: analysis of postoperative clinical outcomes between patients with and without a history of previous cervical spine surgery. World Neurosurg. 2018;109:e144-e149. doi: 10.1016/j.wneu.2017.09.122.

40. Goneidy A, Ong EGP. Effect of previous abdominal surgery and gallbladder appearance on biliary atresia outcomes. J Pediatr Surg. 2018;53(2):274-276. doi: 10.1016/j.jpedsurg.2017.11.018.

41. Chu D, Bakaeen FG, Wang XL, LeMaire SA, Coselli JS, Huh J. Does the duration of surgery affect outcomes in patients undergoing coronary artery bypass grafting? Am J Surg. 2008;196(5):652-656. doi: 10.1016/j.amjsurg.2008.07.003.

42. Kim JY, Khavanin N, Rambachan A, et al. Surgical duration and risk of venous thromboembolism. JAMA Surg. 2015;150(2):110-117. doi: 10.1001/jamasurg.2014.1841.