Readmissions Among Children Previously Hospitalized ... - Pediatrics

Readmissions Among Children Previously Hospitalized With Pneumonia WHAT’S KNOWN ON THIS SUBJECT: Pneumonia is a leading cause of hospitalization among children, and readmissions after discharge are common. WHAT THIS STUDY ADDS: Eight percent of children experience a readmission within 30 days after hospital discharge for pneumonia. Readmissions are most common among young children and those with chronic medical conditions, and are associated with substantial costs.

abstract BACKGROUND AND OBJECTIVES: Pneumonia is a leading cause of hospitalization and readmission in children. Understanding the patient characteristics associated with pneumonia readmissions is necessary to inform interventions to reduce avoidable hospitalizations and related costs. The objective of this study was to characterize readmission rates, and identify factors and costs associated with readmission among children previously hospitalized with pneumonia. METHODS: Retrospective cohort study of children hospitalized with pneumonia at the 43 hospitals included in the Pediatric Health Information System between January 1, 2008, and December 31, 2011. The primary outcome was all-cause readmission within 30 days after hospital discharge, and the secondary outcome was pneumonia-specific readmission. We used multivariable regression models to identify patient and hospital characteristics and costs associated with readmission. RESULTS: A total of 82 566 children were hospitalized with pneumonia (median age, 3 years; interquartile range 1–7). Thirty-day all-cause and pneumonia-specific readmission rates were 7.7% and 3.1%, respectively. Readmission rates were higher among children ,1 year of age, as well as in patients with previous hospitalizations, longer index hospitalizations, and complicated pneumonia. Children with chronic medical conditions were more likely to experience all-cause (odds ratio 3.0; 95% confidence interval 2.8–3.2) and pneumonia-specific readmission (odds ratio 1.8; 95% confidence interval 1.7–2.0) compared with children without chronic medical conditions. The median cost of a readmission ($11 344) was higher than that of an index admission ($4495; P = .01). Readmissions occurred in 8% of pneumonia hospitalizations but accounted for 16.3% of total costs for all pneumonia hospitalizations. CONCLUSIONS: Readmissions are common after hospitalization for pneumonia, especially among young children and those with chronic medical conditions, and are associated with substantial costs. Pediatrics 2014;134:100–109

100

NEUMAN et al

AUTHORS: Mark I. Neuman, MD, MPH,a,b Matthew Hall, PhD,c James C. Gay, MD,d Anne J. Blaschke, MD, PhD,e,f Derek J. Williams, MD, MPH,g,h Kavita Parikh, MD,i,j Adam L. Hersh, MD, PhD,e,g Thomas V. Brogan, MD,k,l Jeffrey S. Gerber, MD, PhD,m,n Carlos G. Grijalva, MD, MPH,o and Samir S. Shah, MD, MSCEp,q aDivision

of Emergency Medicine, Boston Children’s Hospital, Boston, Massachusetts; bDepartment of Pediatrics, Harvard Medical School, Boston, Massachusetts; cThe Children’s Hospital Association, Overland Park, Kansas; Divisions of dGeneral Pediatrics, and hHospital Medicine, Monroe Carell Jr. Children’s Hospital at Vanderbilt, Nashville, Tennessee; Departments of gPediatrics, and oHealth Policy, Vanderbilt University School of Medicine, Nashville, Tennessee; eDivision of Infectious Diseases, Primary Children’s Medical Center, Salt Lake City, Utah; fDepartment of Pediatrics, University of Utah School of Medicine, Salt Lake City, Utah; iDivision of Hospital Medicine, Children’s National Medical Center, Washington, District of Columbia; jDepartment of Pediatrics, George Washington University School of Medicine, Washington, District of Columbia; kDivision of Critical Care, Seattle Children’s Hospital, Seattle, Washington; lDepartment of Pediatrics, University of Washington School of Medicine, Seattle, Washington; mDivision of Infectious Diseases, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania; nDepartment of Pediatrics, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania; pDivisions of Infectious Diseases and Hospital Medicine, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio; and qDepartment of Pediatrics, University of Cincinnati College of Medicine, Cincinnati, Ohio KEY WORDS pneumonia, readmission ABBREVIATIONS CCI—chronic condition indicator CI—confidence interval FPL—federal poverty level ICD-9—International Classification of Diseases, Ninth Revision IQR—interquartile range LOS—length of stay OR—odds ratio PHIS—Pediatric Health Information System

(Continued on last page)

ARTICLE

Hospital readmissions account for a large proportion of health care expenditures.1 Readmission rates have been proposed as a marker of the quality of hospital care, and state and federal agencies have imposed financial penalties on institutions with high rates of readmission.2 Although efforts to reduce readmissions have focused on chronic medical conditions, acute conditions account for a large proportion of readmissions in children.3 Pneumonia is a leading cause of hospitalization among children, and readmissions after discharge are common.3,4 A better understanding of the patient and hospital characteristics associated with readmission among children previously hospitalized with pneumonia may help to target interventions to reduce unnecessary repeat hospitalizations and costs. However, few studies have tried to identify the factors associated with higher rates of pneumonia readmission. A previous study described variability in the 14-day readmission rates for children previously hospitalized with uncomplicated pneumonia in US children’s hospitals (range of 1.5%–4.4%).5 In that study, readmission rates were not associated with diagnostic testing or with hospital length of stay (LOS),5 but other factors were not assessed in detail. The objectives of this study were to (1) characterize the rates of readmission within 30 days after discharge of children hospitalized with pneumonia, (2) compare readmission rates across US children’s hospitals, (3) identify patient- and hospital-level factors associated with readmission after pneumonia hospitalizations, and (4) evaluate the costs of readmissions relative to index admissions.

METHODS We conducted a retrospective cohort study using data from the Pediatric Health Information System (PHIS), which PEDIATRICS Volume 134, Number 1, July 2014

contains administrative and resource utilization data from freestanding children’s hospitals. Participating hospitals are located in noncompeting markets of 27 states plus the District of Columbia and account for ∼15% of all pediatric hospitalizations in the United States.6 Participating hospitals provide patient-level data, including demographic characteristics, diagnoses, and procedures, as well as billing data, which includes all medication, diagnostic imaging, laboratory, and supply charges to individual patients. All-Patient Refined Diagnosis-Related Groups (3M Health Information Systems, Provo, UT) also are assigned to each hospitalization by PHIS using the data supplied by participating hospitals and can be used to describe the principal reason for an admission. Data are de-identified before inclusion in the database; however, encrypted medical record numbers allow for tracking individual patients across hospital visits and admissions. The Children’s Hospital Association (Overland Park, KS) and participating hospitals jointly ensure the quality and integrity of the data, as previously described.7 This study was approved by the institutional review board at Boston Children’s Hospital. Children hospitalized for pneumonia at 1 of the 43 PHIS participating children’s hospitals between January 1, 2008, and December 31, 2011, were included. We defined pneumonia using a previously validated set of International Classification of Diseases, Ninth Revision (ICD-9) discharge diagnosis codes: (1) primary diagnosis of pneumonia (481–483.8, 485–486); or (2) primary diagnosis of pleural effusion (510.0, 510.9, 511.0, 511.1, and 511.9) and a secondary diagnosis of pneumonia.8 This definition had a sensitivity of ∼72% and a specificity of 91% for providerconfirmed community-acquired pneumonia.8 We excluded children who died during their hospitalization, as well as children with cystic fibrosis, in whom

the diagnosis of pneumonia can be difficult to distinguish from an exacerbation of their underlying disease. Outcome Measures Readmission Our primary outcome was all-cause readmission within 30 days after discharge of children hospitalized with pneumonia. Our secondary outcome was pneumonia-specific readmission, defined as a readmission within 30 days with an ICD-9 discharge diagnosis code for pneumonia in any primary or nonprimary position. Costs Costs were calculated for the initial (index) hospitalization, readmission, and combined costs, including the costs of the index hospitalization and any subsequent hospitalization(s) occurring within 30 days of the index hospitalization’s discharge date. All costs were estimated from charges reported to PHIS by using each hospital’s costto-charge ratio adjusted for hospital location by using the Centers for Medicare and Medicaid Services’ price/wage index.9 We calculated the overall costs for all pneumonia hospitalizations and also compared the costs for index hospitalizations with costs for hospitalizations that were readmissions. Exposures Hospital Characteristics Hospitals were characterized based on geographic region (Northeast, South, Midwest, and West),10 the annual number of hospitalizations (by quartile), annual number of pneumonia hospitalizations (by quartile), and payer-mix, based on the proportion using government-based insurance (by quartile). Characteristics of Children Demographic characteristics analyzed included age (0–30 days, 31–365 days, 101

1–4 years, 5–12 years, 13–18 years, or .18 years), gender, race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, Asian, or other), and insurance type (percentage of children with government insurance). We combined commercial and other insurance into a single nongovernment insurance category because commercial insurance accounted for the vast majority of insurance types (∼77%). Nongovernment insurance included patients with commercial insurance as well as those categorized as self-pay, other, or missing. By using 5-digit zip code data, patients also were classified based on the distance of their residence from the hospital (,20 vs $20 miles), and median household income (#1.5 times the federal poverty level [FPL], 1.5–23FPL, .2–33FPL, and .33FPL). The Agency for Healthcare Research and Quality Chronic Condition Indicator (CCI) classification system was used to identify the presence of a chronic medical condition, which categorizes more than 14 000 ICD-9 diagnosis codes into chronic versus nonchronic conditions.11 Categorization of a chronic medical condition used diagnosis codes either during their index encounter or during the readmission, and was categorized as a binary indicator of presence, as well as by the number of CCIs (0, 1, 2, 3, $4) as a surrogate for medical complexity similar to previous studies.3,12,13 Patients were additionally categorized based on the presence of an asthma diagnosis (493.3), as this condition may predispose a child to pneumonia, and because it is often difficult to distinguish asthma from pneumonia based on clinical or radiographic findings. Patients also were classified as requiring technology assistance if medical technology was used to maintain a child’s health status, such as a tracheostomy, cerebrospinal fluid ventricular shunt, or gastrostomy.4 Patients were further characterized based on the number of hospitalizations 102

NEUMAN et al

in the year preceding the index hospitalization (0, 1, 2, $3), whether they were hospitalized for pneumonia in the previous year, the length of their index hospitalization (1–2 days, 3–7 days, or .7 days), and whether they received ICU care. Last, patients were classified based on whether they had an ICD-9 diagnosis code for complicated pneumonia, and whether they underwent a pleural drainage procedure. Complicated pneumonia was defined by the presence of empyema (510.0 and 510.9), pleurisy (511.0 and 511.1), pleural effusion (511.9), or abscess of lung (513.0), and pleural drainage was defined by ICD-9 procedure codes for thoracentesis (34.91), chest tube placement (34.04), video-assisted thoracoscopic surgery (34.21), and thoracotomy (34.02 and 34.09). Statistical Analyses Hospital- and patient-level characteristics were summarized using frequencies and percentages or median (interquartilerange[IQR]) forcategorical and continuous variables, respectively. We derived multivariate logistic regression models to evaluate the independent role of patient- and hospital-level covariates on the odds of experiencing a 30-day readmission. Separate models were derived for all-cause and pneumonia-specific readmission. P , .05 was considered statistically significant for all analyses. All analyses were performed using SAS v.9.2 (SAS Institute, Cary, NC).

RESULTS The study cohort consisted of 82 566 children hospitalized with pneumonia; median age was 3 years (IQR 1–7). Overall, 7.7% (95% confidence interval [CI] 7.5–7.9) of patients experienced an all-cause readmission within 30 days of their index pneumonia hospitalization discharge, and 3.1% (95% CI 2.9– 3.2) experienced a pneumonia-specific

readmission. Across hospitals, the median 30-day hospital-specific all-cause readmission rate was 8.1% (IQR 6.5–9.3), and pneumonia-specific readmission rate was 3.1% (IQR 2.5–3.9) (Fig 1). For allcause readmissions, 16.7% occurred within 3 days after hospital discharge, 34.0% occurred within 1 week, and 58.5% occurred within 2 weeks after discharge (Fig 2). Hospital Characteristics Associated With Readmission Hospitals located in the Northeast and Midwest had higher 30-day all-cause readmission rates than other hospitals (Table 1). Although the annual volume of all hospitalizations was not associated with 30-day readmission rates, hospitals in the upper quartile of pneumonia hospitalizations had 22% lower odds of experiencing a 30-day readmission for any cause compared with hospitals in the lowest quartile after adjustment. Hospital payer-mix was not associated with readmission rates. Patient Characteristics Associated With Readmission Compared with children 1 to 4 years of age, children ,1 year of age were more likely to experience a readmission after pneumonia hospitalization; children ,1 month of age experienced nearly twice the odds of readmission (adjusted odds ratio [OR] 1.96, 95% CI 1.33–2.87) (Table 2). Readmission was not associated with gender, race/ ethnicity, insurance type, proximity to the hospital, or median household income. Children with asthma had lower odds of 30-day readmission than children without asthma, and children with technology dependence had higher odds of experiencing a pneumonia-specific readmission than those without such dependence. Increased number of previous hospitalizations and increased LOS were both

ARTICLE

FIGURE 1 All-cause and pneumonia-specific 30-day readmission rate by hospital. The total height of the bar represents the median hospital all-cause readmission rate.

associated with readmission after hospitalization for pneumonia. Children receiving intensive care were not more likely to experience a readmission after other factors, including LOS, were accounted for in the analyses, although median hospital LOS was longer for these children (7 vs 2 days; P , .001).

Compared with children with uncomplicated pneumonia, children with complicated pneumonia who did not have a pleural drainage procedure were more likely to experience a readmission within 30 days, whereas this association was not observed among those who underwent a drainage procedure.

Children with CCI(s) were more likely to experience an all-cause (OR 3.01, 95% CI 2.82–3.21) and pneumonia-specific readmission (OR 1.84, 95% CI 1.69– 2.01) within 30 days after hospital discharge compared with children without a CCI. Compared with children without a CCI (all-cause 30-day readmission rate

FIGURE 2 Thirty-day readmissions by day (0–30) after hospitalization for pneumonia.

PEDIATRICS Volume 134, Number 1, July 2014

103

TABLE 1 Hospital Characteristics and Their Association with 30-Day Readmission Hospital Characteristic

Geographic region Northeast South Midwest West No. of annual admissions Q1 (,12 337) Q2 (12 337–14 768) Q3 (14 769–18 218) Q4 (.18 218) No. of annual pneumonia admissions Q1 (,290) Q2 (290–456) Q3 (457–606) Q4 (.606) Payer mix (% government) Q1 (,48%) Q2 (48%–56%) Q3 (57%–65%) Q4 (.65%)

Index Admission, n (%)

All-Cause Readmission (n = 6346, 7.69%)

Pneumonia-Specific Readmissiona (n = 2537, 3.07%)

Unadjusted ORs for 30-d Readmission (95% CI)

Adjusted OR for 30-d Readmission (95% CI)


Adjusted ORs for 30-d Readmission (95% CI)

8889 (10.77) 32 941 (39.90) 22 528 (27.28) 18 208 (22.05)

Ref 0.82 (0.75–0.89)** 0.98 (0.90–1.07) 0.87 (0.79–0.95)*

Ref 0.84 (0.73–0.98)* 1.05 (0.90–1.21) 0.83 (0.74–0.93)**

Ref 0.86 (0.75–0.98)** 1.01 (0.88–1.16) 0.93 (0.80–1.07)

Ref 0.87 (0.71–1.06) 1.03 (0.87–1.23) 0.91 (0.77–1.08)

13 913 (16.85) 19 073 (23.10) 21 897 (26.52) 27 683 (33.53)

Ref 1.11 (1.02–1.21) 1.12 (1.03–1.22) 1.16 (1.08–1.26)**

Ref 1.07 (0.94–1.21) 1.10 (0.97–1.25) 1.19 (1.00–1.27)

Ref 1.07 (0.94–1.22) 1.06 (0.93–1.20) 1.11 (0.98–1.25)

Ref 1.05 (0.89–1.24) 1.04 (0.90–1.21) 1.13 (0.90–1.40)

10 941 (13.25) 14 944 (18.10) 23 131 (28.02) 33 550 (40.63)

Ref 1.06 (0.97–1.16) 0.99 (0.91–1.08) 0.90 (0.83–0.97)**

Ref 0.83 (0.71–0.97)* 0.87 (0.77–1.00)* 0.78 (0.66–0.91)*

Ref 1.12 (0.97–1.29) 1.02 (0.90–1.16) 0.93 (0.82–1.06)

Ref 0.95 (0.75–1.21) 0.93 (0.76–1.15) 0.87 (0.69–1.09)

19 779 (23.96) 21 612 (26.18) 19 007 (23.02) 22 168 (26.85)

Ref 0.91 (0.85–0.98)* 1.05 (0.97–1.13) 0.96 (0.89–1.03)

Ref 1.01 (0.89–1.15) 1.10 (0.95–1.28) 1.09 (0.99–1.21)

Ref 0.98 (0.87–1.09) 1.06 (0.95–1.19) 0.96 (0.86–1.07)

Ref 1.10 (0.92–1.32) 1.14 (0.91–1.32) 1.12 (0.94–1.34)

Q1, quartile 1; Q2, quartile 2; Q3, quartile 3; Q4, quartile 4. * P , .05; ** P , .01. a Readmission within 30 days of discharge from index hospitalization with diagnosis of pneumonia in any diagnosis field (ICD-9 codes 481–483.8, 485–486)

= 3.7%), children with a CCI involving the nervous system, digestive system, or circulatory system were more likely to experience an all-cause readmission (17%, 18%, and 19%, respectively; Fig 3). Children with a CCI involving the respiratory system did not have increased odds of either all-cause or pneumoniaspecific readmission (Supplemental Table 5). Approximately 20.4% of children with conditions influencing health status (such as technology dependence) experienced a readmission within 30 days after pneumonia hospitalization. Reasons for Readmission Eight diagnosis categories accounted for .54% of readmissions within 30 days of index hospitalization discharge. Pneumonia was the most common, responsible for 22.6% of all readmissions (Table 3). Costs The total cost for all pneumonia hospitalizations was ∼$1.0 billion (index and readmission costs combined). Of 104

NEUMAN et al

these total costs, readmissions accounted for $163 million (16%), despite accounting for only 7.7% of total pneumonia hospitalizations during the study period (Table 4). The median cost of the index admission increased with increasing number of CCIs (P, trend , .001), and also was higher for children with complicated versus uncomplicated pneumonia. The median cost of a readmission was higher than for an index admission ($11 344 vs $4495, respectively; P = .01); however, the difference in cost was not significant among children with an index hospitalization of .7 days of duration, and for those undergoing a pleural drainage procedure on their index admission (Table 4).

DISCUSSION In this national cohort of children hospitalized with pneumonia at US children’s hospitals, we identified that young age, presence of chronic comorbidities or complicated pneumonia, and care in hospitals with lower volumes of

admissions for pneumonia were associated with an increased risk of readmission. Readmissions usually occurred shortly after hospital discharge, with more than one-third occurring within 1 week. The median cost of a readmission was more than twice that of an index hospitalization, and although only 8% of children hospitalized with pneumonia experienced a readmission, these readmissions accounted for 16% of the total pneumonia hospitalization costs. Pneumonia is the third leading cause of pediatric readmissions, after only seizures and bronchiolitis, and accounts for 3.2% of all 30-day readmissions.3 Given the variability in readmission rates among hospitals in the current study, and the importance the Centers for Medicare and Medicaid Services has placed on readmissions for pneumonia in adults,14 pneumonia is a valid focus of pediatric readmission reduction efforts. More than one-third of readmissions for childhood pneumonia occur within

ARTICLE

TABLE 2 Patient Characteristics and Their Association With 30-Day Readmission Patient Characteristic on Index Visit

Age 0–30 d 31–365 d 1–4 y 5–12 y 13–18 y .18 y Girls Race/Ethnicity Non-Hispanic white Non-Hispanic black Hispanic Asian Other Government payer Proximity to hospital, miles ,20 $20 Median household incomeb #1.53FPL 1.5–23FPL 2–33FPL .33FPL CCI presentc No. of CCIs 0 1 2 3 4 or more Asthma Technology dependence No. hospitalizations in previous y 0 1 2 3+ Pneumonia-specific hospitalization in previous y Hospital LOS, d 1–2 3–7 .7 ICU care received Complicated pneumonia None (uncomplicated) Complicated pneumonia diagnosis aloned Complicated pneumonia with pleural drainage proceduree

Index Admission, n (%)

All-Cause Readmission, n = 6346, 7.69%

Pneumonia-Specific Readmission,a n = 2537, 3.07%





422 (0.51) 12 509 (15.15) 37 391 (45.29) 23 948 (29.00) 6893 (8.35) 1403 (1.70) 38 220 (46.29)

1.24 (0.87–1.77) 1.64 (1.53–1.76)** Ref 0.99 (0.92–1.05) 1.74 (1.59–1.89)** 2.94 (2.55–3.40)** 0.91 (0.86–0.96)*

1.96 (1.33–2.87)** 1.73 (1.63–1.83)** Ref 0.87 (0.82–0.93)** 1.01 (0.91–1.11) 0.97 (0.82–1.15) 0.96 (0.91–1.02)

1.06 (0.60–1.89) 1.25 (1.11–1.40)** Ref 1.12 (1.02–1.24)* 1.69 (1.48–1.92)** 2.28 (1.81–2.88)** 0.92 (0.85–1.00)

1.48 (0.62–3.52) 1.36 (1.14–1.61)** Ref 1.02 (0.92–1.13) 1.18 (1.00–1.39) 1.06 (0.82–1.36) 0.94 (0.85–1.04)

34 769 (45.66) 18 563 (24.38) 19 753 (25.94) 2287 (3.00) 775 (1.02) 47 632 (57.69)

Ref 0.93 (0.87–0.99)* 1.03 (0.96–1.09) 0.77 (0.64–0.91)* 0.98 (0.75–1.29) 1.22 (1.16–1.29)*

Ref 1.02 (0.92–1.12) 1.05 (0.98–1.14) 0.90 (0.73–1.12) 1.16 (0.86–1.57) 0.95 (0.89–1.02)

Ref 0.90 (0.81–1.00) 0.99 (0.89–1.09) 0.85 (0.65–1.10) 1.10 (0.74–1.62) 1.03 (0.95–1.12)

Ref 1.00 (0.85–1.18) 1.05 (0.94–1.16) 0.97 (0.74–1.29) 1.25 (0.89–1.75) 0.88 (0.76–1.02)

55 501 (69.10) 24 819 (30.90)

Ref 1.21 (1.15–1.28)*

Ref 1.02 (0.95–1.09)

Ref 1.07 (0.98–1.17)

Ref 0.94 (0.84–1.06)

22 317 (27.68) 24 799 (30.76) 24 840 (30.81) 8675 (10.76) 49 962 (60.51)

Ref 0.99 (0.92–1.06) 0.93 (0.87–1.00) 0.93 (0.85–1.02) 3.01 (2.82–3.21)**

Ref 0.97 (0.89–1.05) 0.99 (0.91–1.07) 1.03 (0.92–1.15) —

Ref 0.93 (0.84–1.04) 0.99 (0.89–1.10) 0.97 (0.84–1.12) 1.84 (1.69–2.01)**

Ref 0.89 (0.79–1.00) 1.00 (0.88–1.13) 0.95 (0.79–1.14) —

32 604 (39.49) 26 376 (31.95) 9595 (11.62) 5697 (6.90) 8294 (10.05) 25 159 (30.47) 11 737 (14.22)

Ref 1.42 (1.31–1.53)** 3.38 (3.10–3.68)** 5.10 (4.65–5.58)** 7.10 (6.57–7.68)** 0.71 (0.67–0.76)** 4.22 (4.00–4.46)**

Ref 1.58 (1.42–1.75)** 2.41 (2.14–2.72)** 2.86 (2.52–3.25)** 3.38 (2.93–3.89)** 0.62 (0.57–0.68)** 1.07 (0.98–1.16)

Ref 1.09 (0.97–1.21) 1.94 (1.71–2.21)** 2.87 (2.50–3.28)** 3.54 (3.16–3.97)** 0.78 (0.71–0.86)** 2.85 (2.61–3.11)**

Ref 1.19 (1.01–1.39)* 1.52 (1.27–1.82)** 1.88 (1.58–2.25)** 1.97 (1.60–2.42)** 0.74 (0.64–0.85)** 1.15 (1.03–1.29)*

59 179 (71.67) 9131 (11.06) 5845 (7.08) 8411 (10.19) 6243 (7.56)

Ref 2.83 (2.62–3.06)** 3.46 (3.17–3.77)** 8.56 (8.04–9.12)** 3.11 (2.90–3.34)**

Ref 1.92 (1.76–2.09)** 2.31 (2.10–2.54)** 4.60 (4.22–5.01)** 0.95 (0.86–1.03)

Ref 1.83 (1.62–2.07)** 2.09 (1.82–2.40)** 4.55 (4.14–5.00)** 2.98 (2.69–3.31)**

Ref 1.48 (1.32–1.67)** 1.47 (1.24–1.74)** 2.70 (2.35–3.10)** 1.37 (1.17–1.61)**

44 215 (53.55) 28 882 (34.98) 9469 (11.47) 9509 (11.52)

Ref 1.72 (1.63–1.83)** 3.36 (3.13–3.60)** 2.20 (2.06–2.35)**

Ref 1.13 (1.05–1.22)** 1.39 (1.23–1.56)** 1.10 (1.00–1.21)

Ref 1.47 (1.35–1.61)** 2.08 (1.86–2.32)** 1.66 (1.49–1.84)**

Ref 1.05 (0.96–1.16) 0.97 (0.85–1.11) 1.07 (0.96–1.20)

73 247 (88.71) 6457 (7.82)

Ref 1.08 (0.99–1.19)

Ref 1.32 (1.11,1.56)**

Ref 1.62 (1.43–1.83)**

Ref 1.82 (1.44–2.31)**

2862 (3.47)

0.91 (0.79–1.05)

0.99 (0.82–1.20)

1.11 (0.90–1.38)

1.19 (0.89–1.59)

* P , .05. ** P , .01. —, CCI as a binary variable (present vs absent) was not considered as a candidate predictor in the multivariable model, as it was used in the model as an ordinal variable (No. of CCI’s). a Readmission within 30 days of discharge from index hospitalization with diagnosis of pneumonia in any diagnosis field (ICD-9 codes 481–483.8, 485–486). b Median household income based on zip code of residence. c CCIs developed by the Agency for Healthcare Research and Quality categorize more than 14 000 ICD-9 codes into chronic versus nonchronic conditions. d Complicated pneumonia defined by the presence of empyema (510.0 and 510.9), pleurisy (511.0 and 511.1), pleural effusion (511.9), and abscess of lung (513.0). e Pleural drainage procedure defined by ICD-9 codes for thoracentesis (34.91), chest tube placement (34.04), video-assisted thoracoscopic surgery (34.21), and thoracotomy (34.02 and 34.09).


105

FIGURE 3 All-cause and pneumonia-specific 30-day readmission rate by chronic medical condition.

1 week after hospital discharge, a period in which patients are vulnerable to deterioration of their respiratory status from underlying pneumonia and susceptible to other disease conditions. This observation should focus attention on providing anticipatory guidance to patients likely to return, and on improving coordination of care with outpatient providers and reinforcing the importance of follow-up within a few days of discharge. Interventions, such

as comprehensive discharge planning,15 and communication with primary care providers at the time of discharge, have been shown to reduce readmissions after hospitalization in adults.16 Discharge guidelines and clinical pathways have been shown to reduce readmissions after hospitalization for other respiratory conditions, such as bronchiolitis17 and asthma18 in children. Although we observed no association between overall hospital volume and

TABLE 3 Most Common All-Cause Readmission Diagnoses for Children With Readmission Within 30 Days of Index Hospitalization APR-DRG 139 137 141 138 144 130 132 113

Diagnosis

n

%

Pneumonia not elsewhere classified Major respiratory infections and inflammations Asthma Bronchiolitis and respiratory syncytial virus pneumonia Respiratory signs, symptoms, and minor diagnoses Respiratory system diagnosis with ventilator support 96+ h Bronchopulmonary dysplasia and other chronic respiratory disease arising in perinatal period Infections of upper respiratory tract

1437 360 356 334 261 241 237

22.6 5.7 5.6 5.3 4.1 3.8 3.7

223

3.5

Readmission diagnosis based on the All Patient-Refined Diagnosis-Related Groups (APR-DRG) code.

106

NEUMAN et al

pneumonia readmissions, we did find that hospitals with higher volume of pneumonia hospitalizations had lower readmission rates. Similar to the observation that hospitals that perform more surgical procedures have lower rates of surgical readmissions,19–21 our data suggest that the same paradigm might apply to medical conditions, such as pneumonia. It is conceivable that hospitals with more experience with a given condition, be it medical or surgical, may become more adept at dealing with all aspects of care, including discharge planning and condition-specific follow-up care. Infants and children ,1 year of age were nearly twice as likely to experience a readmission compared with older children, a finding that highlights the challenges of managing pneumonia in this age group. Pneumonia in infants ,1 year is often caused by viruses, which have a protracted and waxing and waning course that can make it

ARTICLE

TABLE 4 Hospital Costs Based on Index Admission or Readmission Patient Characteristic on Index Visit

Index admission, n (%)

Median Total Costs,a Index Hospitalization Plus Readmission, $ (IQR)

Median Cost of Index Hospitalization,b $ (IQR)

Number Readmitted, n (%)

Median Cost of Readmission,c $ (IQR)

All patients No. of CCIs 0 1 2 3 $4 Hospital LOS, d 1–2 3–7 .7 Complicated pneumonia None (uncomplicated) Complicated pneumonia diagnosis Complicated pneumonia and pleural drainage procedure

82 566

4701

4495

6346 (7.7)

11 344

32 604 (39.5) 26 376 (31.9) 9595 (11.6) 5697 (6.9) 8294 (10.0)

3500 (2228–6024) 4296 (2705–7524) 6710 (3729–14 265) 10 240 (5147–22 347) 16 779 (7817–38 680)

3417 (2196–5735) 4154 (2660–7056) 6155 (3562–11 838) 8856 (4754–17 822) 13 966 (6832–30 233)

1199 (3.7) 1354 (5.1) 1096 (11.4) 928 (16.3) 1769 (21.3)

5232 (2845–11 752) 6380 (3441–14 670) 10 069 (4712–21 671) 12 812 (5873–28 995) 16 192 (7690–37 307)

44 215 (53.6) 28 882 (35.0) 9469 (11.5)

2912 (2040–4131) 7792 (5452–12 087) 30 994 (20 723–51 934)

2838 (2010–3950) 7478 (5346–11 084) 28 937 (19 802–45 907)

2330 (5.3) 2526 (8.7) 1490 (15.7)

6178 (3283–14 123) 10 740 (5100–22 402) 16 028 (6973–39 887)

73 247 (88.7) 6457 (7.8)

4332 (2611–8316) 8834 (4428–19 610)

4161 (2560–7563) 8307 (4307–18 130)

5613 (7.7) 532 (8.2)

9363 (4354–21 995) 12 798 (5416–25 539)

2862 (3.5)

23 012 (15 248–38 059)

22 469 (15 109–36 809)

201 (7.0)

9694 (4346–23 312)

a

Total combined costs of whole cohort (index and readmission) = $1 003 447 052. Total combined costs for index hospitalizations = $839 813 057. c Total cost for readmissions = $163 633 995, and percentage of cost attributable to readmissions = 16.31%. b

difficult to assess readiness for discharge and predict the need to return. Studies of children with asthma22,23 and a single-center study evaluating preventability of readmissions in children24 also have found higher rates of readmission for children ,1 year, which further emphasizes the need to improve inpatient decision-making in this age group. The cost of readmissions for pneumonia is substantial; repeat hospitalizations account for .16% of all costs for pneumonia in children. The cost of a readmission in our study was higher than for an index hospitalization. Interventions that effectively prevent costly readmissions may have important implications for reducing health care costs. New legislation enacted through the Affordable Care Act allows the Centers for Medicare and Medicaid Services to institute penalties for hospitals in which readmission rates are greater than expected for adults with pneumonia.25 However, there remains debate over the use of readmissions as a measure of hospital care quality, particularly in pediatrics. Although our data show a significant cost associated PEDIATRICS Volume 134, Number 1, July 2014

with readmissions after pneumonia hospitalization, one study found that only 20% of all readmissions in children were likely preventable, and suggested targeting interventions to those patients at higher risk.26 Although this study did not examine other respiratory conditions specifically, they constitute a group in whom effective discharge planning has been shown to reduce readmissions.17,18 Thus, evaluation of similar approaches to reduce readmissions after hospitalizations for pneumonia is warranted. Children with complicated pneumonia (primarily parapneumonic effusions) who did not undergo a pleural drainage procedure were 30% more likely to experience an all-cause readmission, and 80% more likely to experience a pneumonia-specific readmission compared with children without complicated pneumonia. In contrast, children with complicated pneumonia who underwent a pleural drainage procedure had readmission rates similar to children without complicated pneumonia. Additional studies addressing this intriguing observation would be helpful.

Children with chronic medical conditions were more likely to experience a readmission after pneumonia hospitalization, and the rate of readmission increased with increasing number of chronic medical conditions. A singlecenter study of readmissions in children observed that 78% of children experiencing a readmission within 15 days had a chronic condition,24 and the presence of a chronic condition also increased the odds of readmission among children with asthma.27 Similar to a pattern observed in adults,28 we found that a history of previous hospitalizations increased the odds of readmission after discharge for pneumonia. Children with high-complexity illnesses should undergo extensive discharge planning, and safeguards should be enacted to ensure these patients are indeed ready for discharge. This study is subject to several limitations inherent to retrospective studies of this nature. We relied on available clinical and administrative data to assess and account for differences in illness severity at the patient or hospital level. Thus, residual confounding 107

cannot be ruled out. We also were unable to ascertain the necessity of the readmission or the criteria used to decide to readmit among and within participating hospitals. Although our all-cause readmission definition may include hospitalizations not related to the index pneumonia hospitalization, many all-cause readmissions were respiratory-related and several were specifically coded for pneumonia, suggesting that these readmissions were likely related to the index hospitalization. It also is possible we would underestimate rate of readmission if the child was rehospitalized at a different institution. Although disease misclassification is always a concern with administrative data, our pneumonia case-definition used validated criteria with a high

specificity, making false-positives (ie, children with a pneumonia code who do not have pneumonia) uncommon.8 We also were unable to evaluate care provided in the outpatient setting after index hospitalization (ie, care that may have prevented rehospitalization). Outpatient care after hospitalization has been shown to reduce readmissions among patients with sickle cell disease29 and heart failure,30 and those patients lacking timely primary care follow-up are more likely to be readmitted.31 Our study was conducted among children cared for at freestanding pediatric hospitals, and such hospitals care for a higher proportion of children with comorbidities; thus, our findings may not be generalizable to nonchildren’s hospitals.4 Last, we are unable to estimate the baseline

rate of pneumonia hospitalization in our study, which is important considering the rate of hospitalizations may be a strong predictor of readmission.32

children with bacterial meningitis. JAMA. 2008;299(17):2048–2055 Williams DJ, Shah SS, Myers A, et al. Identifying pediatric community-acquired pneumonia hospitalizations: accuracy of administrative billing codes. JAMA Pediatr. 2013;167(9):851–858 Healthcare Cost and Utilization Project. Costto-charge ratio files. Available at: www. hcup-us.ahrq.gov/db/state/costtocharge. jsp. 2014. Accessed April 2, 2014 US Census Bureau. Geographic terms and concepts—census divisions and census regions. Available at: https://www.census. gov/geo/reference/gtc/gtc_census_divreg. html. 2014. Accessed April 2, 2014 Agency for Healthcare Research and Quality. Chronic condition indicator (CCI) for ICD-9-CM. Available at: www.hcup-us.ahrq. gov/toolssoftware/chronic/chronic.jsp. Accessed November 5, 2013 Feudtner C, Christakis DA, Connell FA. Pediatric deaths attributable to complex chronic conditions: a population-based study of Washington State, 1980–1997. Pediatrics. 2000;106(1 pt 2):205–209 Kharbanda AB, Hall M, Shah SS, et al. Variation in resource utilization across a national sample of pediatric emergency departments. J Pediatr. 2013;163(1):230–236

14. Centers for Medicare and Medicaid Services. CMS readmissions reduction program. Available at: www.cms.gov/Medicare/MedicareFee-for-Service-Payment/AcuteInpatientPPS/ Readmissions-Reduction-Program.html. 2014. Accessed March 17, 2014 15. Naylor M, Brooten D, Jones R, LavizzoMourey R, Mezey M, Pauly M. Comprehensive discharge planning for the hospitalized elderly. A randomized clinical trial. Ann Intern Med. 1994;120(12):999–1006 16. Coleman EA, Parry C, Chalmers S, Min SJ. The care transitions intervention: results of a randomized controlled trial. Arch Intern Med. 2006;166(17):1822–1828 17. Cheney J, Barber S, Altamirano L, et al. A clinical pathway for bronchiolitis is effective in reducing readmission rates. J Pediatr. 2005; 147(5):622–626 18. Fassl BA, Nkoy FL, Stone BL, et al. The Joint Commission Children’s Asthma Care quality measures and asthma readmissions. Pediatrics. 2012;130(3):482–491 19. Hornik CP, He X, Jacobs JP, et al. Relative impact of surgeon and center volume on early mortality after the Norwood operation. Ann Thorac Surg. 2012;93(6):1992–1997 20. Pasquali SK, Li JS, Burstein DS, et al. Association of center volume with mortality

CONCLUSIONS Approximately 8% of children experience a readmission within 30 days of pneumonia hospitalization, and children ,1 year of age and those with chronic medical conditions are at higher risk of readmission. The costs associated with pneumonia readmissions are substantial. Our findings may help to target interventions and provide anticipatory guidance for clinicians and for patients at greatest risk of hospital readmission.

REFERENCES 1. Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med. 2009;360(14):1418–1428 2. Joynt KE, Jha AK. Characteristics of hospitals receiving penalties under the Hospital Readmissions Reduction Program. JAMA. 2013;309(4):342–343 3. Berry JG, Toomey SL, Zaslavsky AM, et al. Pediatric readmission prevalence and variability across hospitals [published correction appears in JAMA. 2013;309(10):986]. JAMA. 2013;309(4):372–380 4. Berry JG, Hall DE, Kuo DZ, et al. Hospital utilization and characteristics of patients experiencing recurrent readmissions within children’s hospitals. JAMA. 2011;305(7):682– 690 5. Brogan TV, Hall M, Williams DJ, et al. Variability in processes of care and outcomes among children hospitalized with communityacquired pneumonia. Pediatr Infect Dis J. 2012;31(10):1036–1041 6. Knapp JF, Hall M, Sharma V. Benchmarks for the emergency department care of children with asthma, bronchiolitis, and croup. Pediatr Emerg Care. 2010;26(5):364–369 7. Mongelluzzo J, Mohamad Z, Ten Have TR, Shah SS. Corticosteroids and mortality in

108

NEUMAN et al

8.

9.

10.

11.

12.

13.

ARTICLE

21.

22.

23.

24.

and complications in pediatric heart surgery. Pediatrics. 2012;129(2). Available at: www.pediatrics.org/cgi/content/full/129/2/ e370 Tsai TC, Joynt KE, Orav EJ, Gawande AA, Jha AK. Variation in surgical-readmission rates and quality of hospital care. N Engl J Med. 2013;369(12):1134–1142 Fleming L, Wilson N, Bush A. Difficult to control asthma in children. Curr Opin Allergy Clin Immunol. 2007;7(2):190–195 Wever-Hess J, Hermans J, Kouwenberg JM, Duiverman EJ, Wever AM. Hospital admissions and readmissions for asthma in the age group 0–4 years. Pediatr Pulmonol. 2001;31(1):30–36 Gay JC, Hain PD, Grantham JA, et al. Epidemiology of 15-day readmissions to a children’s hospital. Pediatrics. 2011;127(6).

25.

26.

27.

28.

Available at: www.pediatrics.org/cgi/content/ full/127/6/e1505 Sexton DJ. Editorial commentary: “excess readmissions” for pneumonia: a dilemma with a penalty. Clin Infect Dis. 2013;57(3):368–369 Hain PD, Gay JC, Berutti TW, Whitney GM, Wang W, Saville BR. Preventability of early readmissions at a children’s hospital. Pediatrics. 2013;131(1). Available at: www.pediatrics.org/cgi/content/full/131/1/e171 Kenyon CC, Melvin PR, Chiang VW, et al. Rehospitalization for childhood asthma: timing, variation, and opportunities for intervention. J Pediatr. 2014;164(2):300–305 Shorr AF, Zilberberg MD, Reichley R, et al. Readmission following hospitalization for pneumonia: the impact of pneumonia type and its implication for hospitals. Clin Infect Dis. 2013;57(3):362–367

29. Leschke J, Panepinto JA, Nimmer M, Hoffmann RG, Yan K, Brousseau DC. Outpatient follow-up and rehospitalizations for sickle cell disease patients. Pediatr Blood Cancer. 2012;58(3):406–409 30. 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 31. Misky GJ, Wald HL, Coleman EA. Posthospitalization transitions: examining the effects of timing of primary care provider follow-up. J Hosp Med. 2010;5 (7):392–397 32. Epstein AM, Jha AK, Orav EJ. The relationship between hospital admission rates and rehospitalizations. N Engl J Med. 2011;365 (24):2287–2295

(Continued from first page) Drs Neuman, Gay, Blaschke, Parikh, Hersh, Brogan, Gerber, Williams, Grijalva, and Shah conceptualized and designed the study, and drafted the initial manuscript; Dr Hall conceptualized and designed the study, conducted data analyses, and drafted the initial manuscript; and all authors approved the final manuscript as submitted. www.pediatrics.org/cgi/doi/10.1542/peds.2014-0331 doi:10.1542/peds.2014-0331 Accepted for publication Apr 9, 2014 Address correspondence to Mark I. Neuman, MD, MPH, Division of Emergency Medicine, Boston Children’s Hospital, 300 Longwood Ave, Boston, MA 02115. E-mail: [email protected] PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275). Copyright © 2014 by the American Academy of Pediatrics FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose. FUNDING: No external funding. POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.


109