ORIGINAL ARTICLE
Predictors of Readmission in Orthopaedic Trauma Surgery Michiel G. J. S. Hageman, MD, Jeroen K. J. Bossen, MD, R. Malcolm Smith, MD, and David Ring, MD, PhD
Objectives: This study of patients who had operative treatment of skeletal trauma addresses (1) the association between readmission within 30 days of discharge and comorbidities and (2) differences in factors associated with all-cause readmissions and those because of a surgical adverse event.
Design: Retrospective study. Setting: Tertiary care referral center. Patients: Three thousand four hundred fifty-two operations for skeletal trauma between 2008 and 2012 with comorbidities quantified using the updated Charlson comorbidity index (CCI). Outcome Measurement: Readmission to the hospital within 30 days of surgery and the subset of readmissions because of adverse events related directly to surgery. Results: There was a significant association between readmission within 30 days of surgery and higher CCI (P , 0.001), older age (P , 0.001), and marital status (widowed) (P , 0.001). The factors associated with readmission related to an adverse event were identical. The best multivariable logistic regression models for all-cause 30-day readmission and 30-day readmission related to a surgical adverse event included CCI and older age in both models (odds ratio 1.1, P , 0.01, pseudo R2 = 0.03).
Conclusions: Older patients and patients with greater comorbidity are more likely to be readmitted within 30 days of surgery for musculoskeletal trauma, whether for a surgical adverse event or another reason. The best multivariable models predicted very little of the variability in readmission, which reflects the complexity of readmission and the difficulty reducing the risk to a few specific factors. Key Words: orthopaedic surgery, readmission, adverse event
Level of Evidence: Prognostic Level II. See Instructions for Authors for a complete description of levels of evidence. (J Orthop Trauma 2014;28:e247–e249)
Accepted for publication March 11, 2014. From the Orthopaedic Department, Massachusetts General Hospital, Boston, MA. J. K. J. Bossen is supported by a Dutch faculty grant from the Vrije Universiteit Amsterdam for medical students (small grant). M. G. J. S. Hageman is supported by a Dutch research grants from Marti-Keunig Eckhart Stichting, Anna Foundation, and Spinoza Foundation. The authors report no conflict of interest. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions this article on the journal’s Web site (www.jorthotrauma.com). Reprints: David Ring, MD, PhD, Orthopaedic Department, Yawkey Center, Suite 2100, Massachusetts General Hospital, 55 Fruit St, Boston, MA 02114 (e-mail: [email protected]). Copyright © 2014 by Lippincott Williams & Wilkins
J Orthop Trauma Volume 28, Number 10, October 2014
INTRODUCTION Readmission is used as a measure of quality of health care.1,2 It has been estimated that between 9% and 48% of readmissions may be avoidable.1,3–8 An understanding of the factors associated with readmission after specific types of care may influence patient counseling, decision making, and quality improvement programs. According to the Center for Medicare & Medicaid Services, a readmission is an admission to a hospital within 30 days after discharge from the same or another hospital.9 It is increasingly common to compare readmission rates of a specific hospital with the national average for a set of patients with a comparable condition as a measure of quality health care.9–11 It is expected that new policies will reduce payment for hospitals with excess rates of readmission.12 The most frequently reported reasons for hospital readmission overall are infection, unstable respiratory illness, development of a new problem, and male sex.1,13–17 In previous orthopaedic studies, the rate of readmission within 30 days of discharge ranged from 5% to 20% after total knee arthroplasty. 18 Among patients with an operatively treated hip fracture, 93,036 (14%) were readmitted within 30 days 14,17 and 5% were readmitted after spine surgery. 13 Comorbidity and age are the strongest factors associated with readmission within 30 days of discharge after a hip fracture.17,19,20 Age is also known as a risk factor for adverse events requiring readmission after surgery.21 The Charlson comorbidity index (CCI) has been developed to predict mortality within 1 year of hospital admission. As such, we believe it may be a useful measure of patient infirmity and complexity. Its relation with readmission in orthopaedic surgery merits further investigation.22,23 This study addressed the primary null hypothesis that there is no correlation between the CCI and readmission within 30 days of discharge after operative treatment of skeletal trauma.22 Our secondary study questions addressed the influence of demographics and clinical factors on readmission and differences between all-cause readmissions and those because of an adverse event directly related to surgery (eg, infection or implant failure).
PATIENTS AND METHODS Inclusion Criteria Using an institutional review board–approved protocol, we retrospectively reviewed consecutive adult patients with isolated and multiple fractures treated operatively at a level 1 trauma center between January 2008 and December 2011. www.jorthotrauma.com |
e247
J Orthop Trauma Volume 28, Number 10, October 2014
Hageman et al
Data Aggregation From a list of all readmissions to the hospital within 30 days of surgery, we also identified readmissions for adverse events related directly to the operative fracture treatment. Previous studies have shown that 30 days is a useful time point for identifying unplanned readmissions.24 From the medical record and hospital databases, we retrieved demographic information, International Coding of Diseases (ICD-9) codes for comorbidities, readmission data, and procedure data. For patients with more than one operation, the first was considered the index procedure. The ICD-9, diagnosis-related groups, and National Medical Registration were used to calculate the CCI.22 The CCI has been widely used in studies with large administrative databases to quantify the influence of comorbidities. We used the updated CCI by Quan et al23 based on the hazards ratio of individual comorbidities for mortality within 1 year after hospital admission, including 12 comorbidities with various weightings that together lead to a maximum total score of 24. A higher CCI indicates a higher risk of death within 1 year22 (see Appendix, Supplemental Digital Content 1, http://links.lww.com/BOT/A142).
Statistical Analyses Baseline characteristics of study patients were summarized with frequencies and percentages for categorical variables and mean 6 SD for continuous variables. Factors associated with readmission were sought from demographics and injury characteristics in bivariate analysis. The Student t test and x2 test were used to assess association between unplanned readmission and independent variables, such as sex, ethnicity, race, and marital status. One-way analysis of variance and Pearson rho correlation were used to assess the association between CCI and independent variables, including age and sex. Binary logistic regression was performed to predict significant (P , 0.05) categorical outcomes between the compared groups. To identify predictors independently associated with readmission, a backward stepwise binary logistic regression was used, and variables from the bivariate analysis with a P value ,0.08 were included into the regression model.25 To provide a baseline reference rate of patients with a hip fracture, all adult patients in the cohort with hip fractures were separately analyzed. A post hoc power analysis showed that 3452 subjects with the observed effect size of 0.39 provide 99% power to detect a significant difference using a 2-tailed Student t test, setting alpha at 0.05.
RESULTS A total of 3452 patients had surgery for musculoskeletal trauma between 2008 and 2011. There were 1714 (50%) men and 1738 (50%) women with an average age (6SD) of 59 6 21 years (range 18–104 years). There were 2402 (70%) patients who had internal or external fixation, 555 (16%) who had arthroplasty, and 500 (14%) who had other procedures. The mean (6SD) Charlson score was 1.7 6 2.6 points (range 0–19). One hundred eighty-six patients (5.4%) experienced at least one readmission within 30 days of surgery. Among the readmitted patients, 120 (3.5%) patients experienced
e248
| www.jorthotrauma.com
a surgical adverse event and 66 (1.9%) were readmitted for a nonsurgical adverse event. The most frequent adverse events were infection (n = 52, 43%) and fixation failure (n = 18, 15%) (see Table, Supplemental Digital Content 2, http://links.lww.com/BOT/A206). In the bivariate analysis, there was a significant association between readmission within 30 days of surgery and higher CCI (P , 0.001), older age (P , 0.001), and marital status (widowed) (P , 0.001). Factors associated with readmission related to an adverse event directly related to the surgical procedure included higher CCI (P = 0.0007), older age (P = 0.0004), and marital status (widowed) (P = 0.006). There were no significant differences between readmitted patients with a surgical adverse event compared with those without a surgical adverse event (see Table, Supplemental Digital Content 2, http://links.lww.com/BOT/A206). Sixty-two patients (7%) within the cohort of patients with an operative treatment of a hip fracture experienced at least one readmission within 30 days of surgery. Among the readmitted patients, 41 (4.6%) patients experienced a surgical adverse event and 21 (2.4%) were readmitted for a nonsurgical adverse event. The most frequent adverse events were infection (n = 37, 60% of adverse events) and fixation failure (n = 7, 11% of adverse events) (see Table, Supplemental Digital Content 2, http://links.lww.com/BOT/A206). In bivariate analysis, among patients treated operatively for a hip fracture, readmitted patients with a surgical adverse event were significantly older on average than those without a surgical adverse event. The best multivariable logistic regression models for 30-day readmission and 30-day readmission related to a surgical adverse event included CCI and age in both models (odds ratio 1.1, P , 0.01, pseudo R2 = 0.03) but explained very little of the variation in readmission (see Table, Supplemental Digital Content 3, http://links.lww.com/BOT/A207).
DISCUSSION Older, more infirm patients are more likely to be readmitted within 30 days of surgery for musculoskeletal trauma. This is true for readmissions related and unrelated to surgical adverse events, such as infection or fixation failure. These results are in line with previous studies showing that age and comorbidity are strong predictors for complications within 30 days of surgery for spinal metastases.17,21,26,27 The finding that there is no significant difference between readmitted patients with a surgical adverse event compared with those without a surgical adverse event among patients operatively treated for a hip fracture may be related to an inadequate postdischarge care coordination. Although the associations are statistically very strong, the best multivariable models predicted very little of the variability in readmission, which reflects the complexity of readmission and the difficulty reducing the risk to a few specific factors.28 Our study should be interpreted in light of the fact that the data are from a single center. The Charlson index was constructed from hospital-acquired administrative data (ICD9) rather than specific review of the medial records, although the administrative data are based on expert review of medical records. We found no difference between readmissions Ó 2014 Lippincott Williams & Wilkins
J Orthop Trauma Volume 28, Number 10, October 2014
related and unrelated to surgical adverse events (eg, infection, wound problems, hematoma, fixation failure, etc.). Both age and comorbidity were identified as risk factors for 30-day readmission in previous studies of patients with coronary bypass graft surgery and heart failure and a Medicare population with 470 unique conditions.29,30 Those previous studies also identified higher body mass index and longer hospital stay as factors associated with higher rates of readmission and emergency department visits.29–33 The finding that older age is associated with an increased risk of readmission agrees with other recent studies. Often the problems are related to underlying conditions rather than procedure-specific problems. It has been reported that postdischarge care coordination can be a modifiable factor for reducing the risk of readmission. A well-defined treatment plan after discharge that is coordinated among different outpatient providers and the patient and family may reduce the need for readmission. Future studies are needed to address whether specific postdischarge interventions reduce readmission rates.17 Our statistical models clearly implicate age and comorbidity as the risk of readmission, but the majority of the variation is unaccounted for. Factors such as surgeon, surgery category, and type of adverse event were investigated but were not helpful. Readmission is complex, and even if other important factors are tracked and identified, statistical models may only be able to explain a limited percentage of variation in the risk for readmission. Perhaps, future studies can identify potentially important factors that are not captured in administrative databases and track them prospectively, in the context of operational improvement projects.
ACKNOWLEDGMENT The authors thank Johann Blauth for outstanding research assistantship. REFERENCES 1. Benbassat J, Taragin M. Hospital readmissions as a measure of quality of health care: advantages and limitations. Arch Intern Med. 2000;160: 1074–1081. 2. Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med. 2009;360: 1418–1428. 3. Chaput-Toupin E, Czernichow P, Froment L, et al. Are early unforseen rehospitalizations inevitable? [Article in French]. Rev Epidemiol Sante Publique. 1996;44:221–227. 4. Frankl SE, Breeling JL, Goldman L. Preventability of emergent hospital readmission. Am J Med. 1991;90:667–674. 5. Gautam P, Macduff C, Brown I, et al. Unplanned readmissions of elderly patients. Health Bull (Edinb). 1996;54:449–457. 6. Graham H, Livesley B. Can readmissions to a geriatric medical unit be prevented? Lancet. 1983;1:404–406. 7. Kelly JF, McDowell H, Crawford V, et al. Readmissions to a geriatric medical unit: is prevention possible? Aging (Milano). 1992;4:61–67. 8. Oddone EZ, Weinberger M, Horner M, et al. Classifying general medicine readmissions. Are they preventable? Veterans affairs cooperative studies in health services group on primary care and hospital readmissions. J Gen Intern Med. 1996;11:597–607.
Ó 2014 Lippincott Williams & Wilkins
Predictors of Readmission 9. Readmission Reduction Programm. Available at: http://cmsgov/Medicare/ Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/ReadmissionsReduction-Programhtml/. Accessed January 5, 2012. 10. Adeyemo D, Radley S. Unplanned general surgical re-admissions—how many, which patients and why? Ann R Coll Surg Engl. 2007;89:363–367. 11. Friedman B, Basu J. The rate and cost of hospital readmissions for preventable conditions. Med Care Res Rev. 2004;61:225–240. 12. Kocher RP, Adashi EY. Hospital readmissions and the Affordable Care Act: paying for coordinated quality care. JAMA. 2011;306:1794–1795. 13. Amin BY, Tu TH, Schairer WW, et al. Pitfalls of calculating hospital readmission rates based on nonvalidated administrative data sets. J Neurosurg Spine. 2013;18:134–138. 14. Teixeira A, Trinquart L, Raphael M, et al. Outcomes in older patients after surgical treatment for hip fracture: a new approach to characterise the link between readmissions and the surgical stay. Age Ageing. 2009;38:584–589. 15. Chu CM, Chan VL, Lin AW, et al. Readmission rates and life threatening events in COPD survivors treated with non-invasive ventilation for acute hypercapnic respiratory failure. Thorax. 2004;59:1020–1025. 16. Bjorgul K, Novicoff WM, Saleh KJ. Evaluating comorbidities in total hip and knee arthroplasty: available instruments. J Orthop Traumatol. 2010; 11:203–209. 17. Kocher KE, Nallamothu BK, Birkmeyer JD, et al. Emergency department visits after surgery are common for Medicare patients, suggesting opportunities to improve care. Health Aff (Millwood). 2013;32:1600–1607. 18. Vorhies JS, Wang Y, Herndon J, et al. Readmission and length of stay after total hip arthroplasty in a national Medicare sample. J Arthroplasty. 2011;26:119–123. 19. French DD, Bass E, Bradham DD, et al. Rehospitalization after hip fracture: predictors and prognosis from a national veterans study. J Am Geriatr Soc. 2008;56:705–710. 20. Riggs RV, Roberts PS, Aronow H, et al. Joint replacement and hip fracture readmission rates: impact of discharge destination. PM R. 2010;2:806–810. 21. Rebasa P, Mora L, Luna A, et al. Continuous monitoring of adverse events: influence on the quality of care and the incidence of errors in general surgery. World J Surg. 2009;33:191–198. 22. Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373–383. 23. Quan H, Li B, Couris CM, et al. Updating and validating the Charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol. 2011;173:676–682. 24. Chan FW, Wong FY, Yam CH, et al. Risk factors of hospitalization and readmission of patients with COPD in Hong Kong population: analysis of hospital admission records. BMC Health Serv Res. 2011;11:186. 25. Duncan DB. Estimation of the probability of an event as a fucntion of several independent variables. Biometrika. 1967;54:167–179. 26. Arrigo RT, Kalanithi P, Cheng I, et al. Charlson score is a robust predictor of 30-day complications following spinal metastasis surgery. Spine (Phila Pa 1976). 2011;36:E1274–E1280. 27. Davis P, Lay-Yee R, Briant R, et al. Adverse events in New Zealand public hospitals I: occurrence and impact. N Z Med J. 2002;115:U271. 28. Kassin MT, Owen RM, Perez SD, et al. Risk factors for 30-day hospital readmission among general surgery patients. J Am Coll Surg. 2012;215: 322–330. 29. Hannan EL, Zhong Y, Lahey SJ, et al. 30-day readmissions after coronary artery bypass graft surgery in New York State. JACC Cardiovasc Interv. 2011;4:569–576. 30. Vashi AA, Fox JP, Carr BG, et al. Use of hospital-based acute care among patients recently discharged from the hospital. JAMA. 2013; 309:364–371. 31. Brock J, Mitchell J, Irby K, et al. Association between quality improvement for care transitions in communities and rehospitalizations among Medicare beneficiaries. JAMA. 2013;309:381–391. 32. Williams MV. A requirement to reduce readmissions: take care of the patient, not just the disease. JAMA. 2013;309:394–396. 33. Bueno H, Ross JS, Wang Y, et al. Trends in length of stay and short-term outcomes among Medicare patients hospitalized for heart failure, 19932006. JAMA. 2010;303:2141–2147.
www.jorthotrauma.com |
e249
Predictors of Readmission in Orthopaedic Trauma Surgery Michiel G. J. S. Hageman, MD, Jeroen K. J. Bossen, MD, R. Malcolm Smith, MD, and David Ring, MD, PhD
Objectives: This study of patients who had operative treatment of skeletal trauma addresses (1) the association between readmission within 30 days of discharge and comorbidities and (2) differences in factors associated with all-cause readmissions and those because of a surgical adverse event.
Design: Retrospective study. Setting: Tertiary care referral center. Patients: Three thousand four hundred fifty-two operations for skeletal trauma between 2008 and 2012 with comorbidities quantified using the updated Charlson comorbidity index (CCI). Outcome Measurement: Readmission to the hospital within 30 days of surgery and the subset of readmissions because of adverse events related directly to surgery. Results: There was a significant association between readmission within 30 days of surgery and higher CCI (P , 0.001), older age (P , 0.001), and marital status (widowed) (P , 0.001). The factors associated with readmission related to an adverse event were identical. The best multivariable logistic regression models for all-cause 30-day readmission and 30-day readmission related to a surgical adverse event included CCI and older age in both models (odds ratio 1.1, P , 0.01, pseudo R2 = 0.03).
Conclusions: Older patients and patients with greater comorbidity are more likely to be readmitted within 30 days of surgery for musculoskeletal trauma, whether for a surgical adverse event or another reason. The best multivariable models predicted very little of the variability in readmission, which reflects the complexity of readmission and the difficulty reducing the risk to a few specific factors. Key Words: orthopaedic surgery, readmission, adverse event
Level of Evidence: Prognostic Level II. See Instructions for Authors for a complete description of levels of evidence. (J Orthop Trauma 2014;28:e247–e249)
Accepted for publication March 11, 2014. From the Orthopaedic Department, Massachusetts General Hospital, Boston, MA. J. K. J. Bossen is supported by a Dutch faculty grant from the Vrije Universiteit Amsterdam for medical students (small grant). M. G. J. S. Hageman is supported by a Dutch research grants from Marti-Keunig Eckhart Stichting, Anna Foundation, and Spinoza Foundation. The authors report no conflict of interest. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions this article on the journal’s Web site (www.jorthotrauma.com). Reprints: David Ring, MD, PhD, Orthopaedic Department, Yawkey Center, Suite 2100, Massachusetts General Hospital, 55 Fruit St, Boston, MA 02114 (e-mail: [email protected]). Copyright © 2014 by Lippincott Williams & Wilkins
J Orthop Trauma Volume 28, Number 10, October 2014
INTRODUCTION Readmission is used as a measure of quality of health care.1,2 It has been estimated that between 9% and 48% of readmissions may be avoidable.1,3–8 An understanding of the factors associated with readmission after specific types of care may influence patient counseling, decision making, and quality improvement programs. According to the Center for Medicare & Medicaid Services, a readmission is an admission to a hospital within 30 days after discharge from the same or another hospital.9 It is increasingly common to compare readmission rates of a specific hospital with the national average for a set of patients with a comparable condition as a measure of quality health care.9–11 It is expected that new policies will reduce payment for hospitals with excess rates of readmission.12 The most frequently reported reasons for hospital readmission overall are infection, unstable respiratory illness, development of a new problem, and male sex.1,13–17 In previous orthopaedic studies, the rate of readmission within 30 days of discharge ranged from 5% to 20% after total knee arthroplasty. 18 Among patients with an operatively treated hip fracture, 93,036 (14%) were readmitted within 30 days 14,17 and 5% were readmitted after spine surgery. 13 Comorbidity and age are the strongest factors associated with readmission within 30 days of discharge after a hip fracture.17,19,20 Age is also known as a risk factor for adverse events requiring readmission after surgery.21 The Charlson comorbidity index (CCI) has been developed to predict mortality within 1 year of hospital admission. As such, we believe it may be a useful measure of patient infirmity and complexity. Its relation with readmission in orthopaedic surgery merits further investigation.22,23 This study addressed the primary null hypothesis that there is no correlation between the CCI and readmission within 30 days of discharge after operative treatment of skeletal trauma.22 Our secondary study questions addressed the influence of demographics and clinical factors on readmission and differences between all-cause readmissions and those because of an adverse event directly related to surgery (eg, infection or implant failure).
PATIENTS AND METHODS Inclusion Criteria Using an institutional review board–approved protocol, we retrospectively reviewed consecutive adult patients with isolated and multiple fractures treated operatively at a level 1 trauma center between January 2008 and December 2011. www.jorthotrauma.com |
e247
J Orthop Trauma Volume 28, Number 10, October 2014
Hageman et al
Data Aggregation From a list of all readmissions to the hospital within 30 days of surgery, we also identified readmissions for adverse events related directly to the operative fracture treatment. Previous studies have shown that 30 days is a useful time point for identifying unplanned readmissions.24 From the medical record and hospital databases, we retrieved demographic information, International Coding of Diseases (ICD-9) codes for comorbidities, readmission data, and procedure data. For patients with more than one operation, the first was considered the index procedure. The ICD-9, diagnosis-related groups, and National Medical Registration were used to calculate the CCI.22 The CCI has been widely used in studies with large administrative databases to quantify the influence of comorbidities. We used the updated CCI by Quan et al23 based on the hazards ratio of individual comorbidities for mortality within 1 year after hospital admission, including 12 comorbidities with various weightings that together lead to a maximum total score of 24. A higher CCI indicates a higher risk of death within 1 year22 (see Appendix, Supplemental Digital Content 1, http://links.lww.com/BOT/A142).
Statistical Analyses Baseline characteristics of study patients were summarized with frequencies and percentages for categorical variables and mean 6 SD for continuous variables. Factors associated with readmission were sought from demographics and injury characteristics in bivariate analysis. The Student t test and x2 test were used to assess association between unplanned readmission and independent variables, such as sex, ethnicity, race, and marital status. One-way analysis of variance and Pearson rho correlation were used to assess the association between CCI and independent variables, including age and sex. Binary logistic regression was performed to predict significant (P , 0.05) categorical outcomes between the compared groups. To identify predictors independently associated with readmission, a backward stepwise binary logistic regression was used, and variables from the bivariate analysis with a P value ,0.08 were included into the regression model.25 To provide a baseline reference rate of patients with a hip fracture, all adult patients in the cohort with hip fractures were separately analyzed. A post hoc power analysis showed that 3452 subjects with the observed effect size of 0.39 provide 99% power to detect a significant difference using a 2-tailed Student t test, setting alpha at 0.05.
RESULTS A total of 3452 patients had surgery for musculoskeletal trauma between 2008 and 2011. There were 1714 (50%) men and 1738 (50%) women with an average age (6SD) of 59 6 21 years (range 18–104 years). There were 2402 (70%) patients who had internal or external fixation, 555 (16%) who had arthroplasty, and 500 (14%) who had other procedures. The mean (6SD) Charlson score was 1.7 6 2.6 points (range 0–19). One hundred eighty-six patients (5.4%) experienced at least one readmission within 30 days of surgery. Among the readmitted patients, 120 (3.5%) patients experienced
e248
| www.jorthotrauma.com
a surgical adverse event and 66 (1.9%) were readmitted for a nonsurgical adverse event. The most frequent adverse events were infection (n = 52, 43%) and fixation failure (n = 18, 15%) (see Table, Supplemental Digital Content 2, http://links.lww.com/BOT/A206). In the bivariate analysis, there was a significant association between readmission within 30 days of surgery and higher CCI (P , 0.001), older age (P , 0.001), and marital status (widowed) (P , 0.001). Factors associated with readmission related to an adverse event directly related to the surgical procedure included higher CCI (P = 0.0007), older age (P = 0.0004), and marital status (widowed) (P = 0.006). There were no significant differences between readmitted patients with a surgical adverse event compared with those without a surgical adverse event (see Table, Supplemental Digital Content 2, http://links.lww.com/BOT/A206). Sixty-two patients (7%) within the cohort of patients with an operative treatment of a hip fracture experienced at least one readmission within 30 days of surgery. Among the readmitted patients, 41 (4.6%) patients experienced a surgical adverse event and 21 (2.4%) were readmitted for a nonsurgical adverse event. The most frequent adverse events were infection (n = 37, 60% of adverse events) and fixation failure (n = 7, 11% of adverse events) (see Table, Supplemental Digital Content 2, http://links.lww.com/BOT/A206). In bivariate analysis, among patients treated operatively for a hip fracture, readmitted patients with a surgical adverse event were significantly older on average than those without a surgical adverse event. The best multivariable logistic regression models for 30-day readmission and 30-day readmission related to a surgical adverse event included CCI and age in both models (odds ratio 1.1, P , 0.01, pseudo R2 = 0.03) but explained very little of the variation in readmission (see Table, Supplemental Digital Content 3, http://links.lww.com/BOT/A207).
DISCUSSION Older, more infirm patients are more likely to be readmitted within 30 days of surgery for musculoskeletal trauma. This is true for readmissions related and unrelated to surgical adverse events, such as infection or fixation failure. These results are in line with previous studies showing that age and comorbidity are strong predictors for complications within 30 days of surgery for spinal metastases.17,21,26,27 The finding that there is no significant difference between readmitted patients with a surgical adverse event compared with those without a surgical adverse event among patients operatively treated for a hip fracture may be related to an inadequate postdischarge care coordination. Although the associations are statistically very strong, the best multivariable models predicted very little of the variability in readmission, which reflects the complexity of readmission and the difficulty reducing the risk to a few specific factors.28 Our study should be interpreted in light of the fact that the data are from a single center. The Charlson index was constructed from hospital-acquired administrative data (ICD9) rather than specific review of the medial records, although the administrative data are based on expert review of medical records. We found no difference between readmissions Ó 2014 Lippincott Williams & Wilkins
J Orthop Trauma Volume 28, Number 10, October 2014
related and unrelated to surgical adverse events (eg, infection, wound problems, hematoma, fixation failure, etc.). Both age and comorbidity were identified as risk factors for 30-day readmission in previous studies of patients with coronary bypass graft surgery and heart failure and a Medicare population with 470 unique conditions.29,30 Those previous studies also identified higher body mass index and longer hospital stay as factors associated with higher rates of readmission and emergency department visits.29–33 The finding that older age is associated with an increased risk of readmission agrees with other recent studies. Often the problems are related to underlying conditions rather than procedure-specific problems. It has been reported that postdischarge care coordination can be a modifiable factor for reducing the risk of readmission. A well-defined treatment plan after discharge that is coordinated among different outpatient providers and the patient and family may reduce the need for readmission. Future studies are needed to address whether specific postdischarge interventions reduce readmission rates.17 Our statistical models clearly implicate age and comorbidity as the risk of readmission, but the majority of the variation is unaccounted for. Factors such as surgeon, surgery category, and type of adverse event were investigated but were not helpful. Readmission is complex, and even if other important factors are tracked and identified, statistical models may only be able to explain a limited percentage of variation in the risk for readmission. Perhaps, future studies can identify potentially important factors that are not captured in administrative databases and track them prospectively, in the context of operational improvement projects.
ACKNOWLEDGMENT The authors thank Johann Blauth for outstanding research assistantship. REFERENCES 1. Benbassat J, Taragin M. Hospital readmissions as a measure of quality of health care: advantages and limitations. Arch Intern Med. 2000;160: 1074–1081. 2. Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med. 2009;360: 1418–1428. 3. Chaput-Toupin E, Czernichow P, Froment L, et al. Are early unforseen rehospitalizations inevitable? [Article in French]. Rev Epidemiol Sante Publique. 1996;44:221–227. 4. Frankl SE, Breeling JL, Goldman L. Preventability of emergent hospital readmission. Am J Med. 1991;90:667–674. 5. Gautam P, Macduff C, Brown I, et al. Unplanned readmissions of elderly patients. Health Bull (Edinb). 1996;54:449–457. 6. Graham H, Livesley B. Can readmissions to a geriatric medical unit be prevented? Lancet. 1983;1:404–406. 7. Kelly JF, McDowell H, Crawford V, et al. Readmissions to a geriatric medical unit: is prevention possible? Aging (Milano). 1992;4:61–67. 8. Oddone EZ, Weinberger M, Horner M, et al. Classifying general medicine readmissions. Are they preventable? Veterans affairs cooperative studies in health services group on primary care and hospital readmissions. J Gen Intern Med. 1996;11:597–607.
Ó 2014 Lippincott Williams & Wilkins
Predictors of Readmission 9. Readmission Reduction Programm. Available at: http://cmsgov/Medicare/ Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/ReadmissionsReduction-Programhtml/. Accessed January 5, 2012. 10. Adeyemo D, Radley S. Unplanned general surgical re-admissions—how many, which patients and why? Ann R Coll Surg Engl. 2007;89:363–367. 11. Friedman B, Basu J. The rate and cost of hospital readmissions for preventable conditions. Med Care Res Rev. 2004;61:225–240. 12. Kocher RP, Adashi EY. Hospital readmissions and the Affordable Care Act: paying for coordinated quality care. JAMA. 2011;306:1794–1795. 13. Amin BY, Tu TH, Schairer WW, et al. Pitfalls of calculating hospital readmission rates based on nonvalidated administrative data sets. J Neurosurg Spine. 2013;18:134–138. 14. Teixeira A, Trinquart L, Raphael M, et al. Outcomes in older patients after surgical treatment for hip fracture: a new approach to characterise the link between readmissions and the surgical stay. Age Ageing. 2009;38:584–589. 15. Chu CM, Chan VL, Lin AW, et al. Readmission rates and life threatening events in COPD survivors treated with non-invasive ventilation for acute hypercapnic respiratory failure. Thorax. 2004;59:1020–1025. 16. Bjorgul K, Novicoff WM, Saleh KJ. Evaluating comorbidities in total hip and knee arthroplasty: available instruments. J Orthop Traumatol. 2010; 11:203–209. 17. Kocher KE, Nallamothu BK, Birkmeyer JD, et al. Emergency department visits after surgery are common for Medicare patients, suggesting opportunities to improve care. Health Aff (Millwood). 2013;32:1600–1607. 18. Vorhies JS, Wang Y, Herndon J, et al. Readmission and length of stay after total hip arthroplasty in a national Medicare sample. J Arthroplasty. 2011;26:119–123. 19. French DD, Bass E, Bradham DD, et al. Rehospitalization after hip fracture: predictors and prognosis from a national veterans study. J Am Geriatr Soc. 2008;56:705–710. 20. Riggs RV, Roberts PS, Aronow H, et al. Joint replacement and hip fracture readmission rates: impact of discharge destination. PM R. 2010;2:806–810. 21. Rebasa P, Mora L, Luna A, et al. Continuous monitoring of adverse events: influence on the quality of care and the incidence of errors in general surgery. World J Surg. 2009;33:191–198. 22. Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373–383. 23. Quan H, Li B, Couris CM, et al. Updating and validating the Charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol. 2011;173:676–682. 24. Chan FW, Wong FY, Yam CH, et al. Risk factors of hospitalization and readmission of patients with COPD in Hong Kong population: analysis of hospital admission records. BMC Health Serv Res. 2011;11:186. 25. Duncan DB. Estimation of the probability of an event as a fucntion of several independent variables. Biometrika. 1967;54:167–179. 26. Arrigo RT, Kalanithi P, Cheng I, et al. Charlson score is a robust predictor of 30-day complications following spinal metastasis surgery. Spine (Phila Pa 1976). 2011;36:E1274–E1280. 27. Davis P, Lay-Yee R, Briant R, et al. Adverse events in New Zealand public hospitals I: occurrence and impact. N Z Med J. 2002;115:U271. 28. Kassin MT, Owen RM, Perez SD, et al. Risk factors for 30-day hospital readmission among general surgery patients. J Am Coll Surg. 2012;215: 322–330. 29. Hannan EL, Zhong Y, Lahey SJ, et al. 30-day readmissions after coronary artery bypass graft surgery in New York State. JACC Cardiovasc Interv. 2011;4:569–576. 30. Vashi AA, Fox JP, Carr BG, et al. Use of hospital-based acute care among patients recently discharged from the hospital. JAMA. 2013; 309:364–371. 31. Brock J, Mitchell J, Irby K, et al. Association between quality improvement for care transitions in communities and rehospitalizations among Medicare beneficiaries. JAMA. 2013;309:381–391. 32. Williams MV. A requirement to reduce readmissions: take care of the patient, not just the disease. JAMA. 2013;309:394–396. 33. Bueno H, Ross JS, Wang Y, et al. Trends in length of stay and short-term outcomes among Medicare patients hospitalized for heart failure, 19932006. JAMA. 2010;303:2141–2147.
www.jorthotrauma.com |
e249
