Comments
Abstract
Rationale: Efforts to reduce 30-day readmission have mostly concentrated on addressing deficiencies in care transitions and outpatient management after discharge. There is growing evidence to suggest that physical inactivity is associated with increased hospitalizations.
Objectives: We examined whether or not a potentially modifiable factor such as regular physical activity at baseline was associated with lower risk of 30-day readmission in patients with chronic obstructive pulmonary disease (COPD).
Methods: Patients from a large integrated health system were included in this retrospective cohort study if they were hospitalized for COPD (following the Centers for Medicare and Medicaid Services and National Quality Forum proposed criteria) and discharged between January 1, 2011 and December 31, 2012, aged 40 years or older, on a bronchodilator or steroid inhaler, alive at discharge, and continuously enrolled in the health plan 12 months before the index admission and at least 30 days post discharge. Our main outcome was 30-day all-cause readmission. Regular physical activity was routinely assessed at the time of all outpatient visits and expressed as the total minutes of moderate or vigorous physical activity (MVPA) per week.
Measurements and Main Results: The sample included a total of 4,596 patients (5,862 index admissions) with a mean age of 72.3 ± 11 years. The 30-day readmission rate was 18%, with 59% of readmissions occurring in the first 15 days. Multivariate adjusted analyses showed that patients reporting any level of MPVA had a significantly lower risk of 30-day readmission compared with inactive patients (1–149 min/wk of MVPA: relative risk, 0.67; 95% confidence interval, 0.55–0.81; ≥150 min/wk of MVPA: relative risk, 0.66; 95% confidence interval, 0.51–0.87). Other significant independent predictors of increased readmission included anemia, prior hospitalizations, longer lengths of stay, more comorbidities, receipt of a new oxygen prescription at discharge, use of the emergency department or observational stay before the readmission (all, P < 0.05), and being unpartnered (P = 0.08).
Conclusions: Our findings further support the importance of physical activity in the management of COPD across the care continuum. Although it is possible that lower physical activity is a reflection of worse disease, promoting and supporting physical activity is a promising strategy to reduce the risk of readmission.
The Centers for Medicare and Medicaid Services began publicly reporting 30-day risk-standardized readmission rates for heart failure, acute myocardial infarction, and pneumonia and penalizing hospitals for having higher than expected all-cause readmission rates in late 2012 (1, 2). Reducing 30-day hospital readmissions has become a major focus of many health care systems, with most efforts concentrating heavily on addressing deficiencies in care transitions and outpatient management after discharge (3–5). In 2015, the list of conditions will expand to include patients admitted for an exacerbation of chronic obstructive pulmonary disease (COPD) (1).
COPD is a chronic progressive disease that is characterized by exacerbations requiring a change in medication or more intensive treatment (6). A number of these episodes are severe enough to require admission to the hospital, with such hospitalizations accounting for nearly 75% of the total direct health care costs for COPD (7). Patients who are hospitalized for a severe COPD exacerbation are at high risk for readmission and death (8–10). National statistics from 2008 show that the 30-day COPD-specific and all-cause readmission rates after an index COPD admission were 7 and 21%, respectively (11). Compared with the index hospitalization, readmission for COPD as the principal diagnosis cost 18% more, whereas costs due to readmissions for all other causes cost over 50% more than the index hospital stay (11). Nearly two-thirds of readmissions from an index COPD admission are due to non-COPD causes (12).
Numerous readmission risk prediction models have been developed to identify which patients might benefit most from care transition interventions and to risk-adjust readmission rates for the purposes of comparisons. Most of these models incorporate variables for medical comorbidity and previous health resource use, with few that include potentially modifiable variables, such as physical functioning, that could be a target for intervention (13). Given the mounting evidence linking physical inactivity to increased hospitalizations and mortality in COPD (14–19), we examined whether regular physical activity captured as part of routine clinical care before the index admission was associated with lower risk of 30-day readmission in COPD. Some of the results from this study have been previously reported in the form of an abstract (20).
This was a retrospective cohort study of members from a large integrated health care system with 14 medical centers located in southern California. Human subject approval was obtained from the Kaiser Permanente Institutional Review Board. Patients were selected for inclusion if they were hospitalized for COPD and discharged between January 1, 2011 and December 31, 2012. COPD hospitalizations were defined according to the Centers for Medicare and Medicaid Services and National Quality Forum proposed criteria: International Classification of Diseases, Ninth Revision principal diagnosis of COPD (491.21, 491.22, 491.8, 491.9, 492.8, 493.20, 493.21, 493.22, and 496) or principal diagnosis of respiratory failure (518.81, 518.82, 518.84, 799.1) and a secondary diagnosis of acute exacerbation of COPD (491.21, 491.22, 493.21, 493.22) and age 40 years or older (21). We applied additional inclusion criteria, including being on at least one maintenance bronchodilator or steroid inhaler in the 12 months before the index admission, alive at discharge, and continuously enrolled in the health plan in the 12 months before index admission and at least 30 days post discharge. Patients who were cared for in an observational stay unit (<24 h of care) were excluded from this analysis. An individual could contribute multiple index admissions during this observation period.
Thirty-day readmission was defined as any repeat admission within 30 days after being discharged alive from an index hospital stay between January 1, 2011 and December 31, 2012. For patients whose hospitalization included transfers between hospitals or within the same hospital, the hospitalization was counted as one event and not counted as a readmission. The resultant length of stay and discharge diagnoses were combined from the individual stays. All readmissions to both Kaiser and non-Kaiser facilities were captured.
During the intake process for all outpatient visits, every patient in our health system is asked the following two questions regarding their regular physical activity (exercise vital sign [EVS]): (1) “On average, how many days per week do you engage in moderate to strenuous (vigorous) exercise (like a brisk walk)?” and (2) “On average, how many minutes do you engage in exercise at this level?” These questions are typically asked by medical assistants and licensed vocational nurses, and patients’ responses are entered into the electronic medical record. Response choices for days are categorical (0–7). Minutes are recorded categorically: 0, 10, 20, 30, 40, 50, 60, 90, 120, and 150 or greater. The electronic medical record system software then multiplies the two self-reported responses to display total minutes per week of moderate or vigorous physical activity (MVPA) for the health care provider to review; we used this calculated value as the exposure variable for our analysis. The EVS closest to the index admission date was used for our primary analysis. Approximately 97% of the EVS data included in the analysis were obtained within 12 months before the index admission (median of 71 d). Due to the skewed MVPA data, we categorized patients as being completely inactive (0 min/wk), insufficiently active (1–149 min/wk), or active, meeting national physical activity recommendations (≥150 min/wk). The EVS has evidence of construct validity in the general health plan population (22, 23).
We extracted administrative and clinical data in the 12 months (January 1, 2010 through December 31, 2011) before the index admission for the entire sample.
Sociodemographics were obtained from our membership files and included age, sex, marital status, race, and insurance status. Body mass index was obtained from the most recent outpatient encounter before the index admission.
Disease severity was based on available spirometry data, and post-bronchodilator FEV1% predicted was used. Only 54% of the sample had any spirometry data available for analysis. A Charlson comorbidity score was calculated for each patient. We also included additional morbidities reportedly associated with hospitalization risk in COPD but not included in the Charlson index, including anemia, anxiety, depression, pulmonary hypertension, and hypertension.
Health resource use included hospitalizations in the 12 months preceding the index event, length of stay of the index event, emergency department or observational stay within 30 days of the index admission but occurring before the readmission event, and discharge disposition at the index admission. Use of inhalers and oxygen supplementation was based on pharmacy and durable medical equipment data from the 12 months before the index admission; to distinguish long-term systemic corticosteroid use versus short-term use to treat COPD exacerbations, patients had to have at least 90 days of continuous steroid use with a refill gap of less than14 days. New oxygen use after the index discharge was based on durable medical equipment data indicating date of oxygen delivery, and only those delivered within the 30 days of discharge were included. Receipt of inpatient palliative care consult during the index admission was also included.
Health behaviors included smoking status, vaccination history, and participation in pulmonary rehabilitation. Smoking status was obtained from the most recent outpatient encounter before their index admission. We were not able to obtain smoking status at the index admission. Vaccination history included influenza and pneumococcal vaccinations in the previous 1 to 2 years and up to 5 years, respectively. Participation in pulmonary rehabilitation up to 4 years before the index admission was recorded.
Bivariate analyses of potential risk factors for 30-day readmission were performed using t tests for continuous variables and Chi-square tests for categorical variables. Significant predictors from the bivariate analyses were included in the final model to test the independent effect of each risk factor using generalized estimating equations with an exchangeable covariance structure to account for patients contributing more than one index admission. Covariates were included in the final model if they contributed substantially to model fit or were of clinical significance. Sensitivity analyses were performed to determine if the effects of physical activity were similar when we used a composite EVS measure of all available values (median or mode) in the 12 months before the index admission, excluded patients with a primary diagnosis of acute respiratory failure (518.8x), and analyzed a subgroup of patients who had spirometry data (54% of the larger sample) to further adjust for disease severity. All significance levels for the generalized estimating equations models were two-sided with a P value less than 0.05. Analyses were conducted using SAS 9.2 (SAS Institute Inc).
We identified 6,042 patients with COPD as their principal discharge diagnosis or who had acute respiratory failure with a secondary diagnosis of acute COPD exacerbation (Figure 1). Patients were excluded if they were less than 40 years old (n = 356), had not been prescribed either a bronchodilator or steroid inhaler before the index admission (n = 905), died during the index stay (n = 139), or had an observational stay (n = 46). The final analytical sample included 4,596 patients who had 5,862 index admissions. Approximately 19% of the patients had more than one index admission during the study period.
Figure 1. Sample flow chart. AECOPD = acute exacerbation of chronic obstructive pulmonary disease; COPD = chronic obstructive pulmonary disease; ICD-9 = International Classification of Diseases, Ninth Revision.
[More] [Minimize]The sample had a mean age of 72.3 ± 11 years and was composed of 68% whites, 12% Hispanics, 4% Asians/Pacific Islanders, and 15% Blacks, with a comparable distribution of men and women (Tables 1 and 2). Approximately 23% had at least one COPD-related hospitalization in the previous 12 months. The mean FEV1% predicted was 57.4 ± 22% for 2,489 patients who had a spirometry test (54% of the sample). Comorbid conditions were common, with a mean Charlson index score of 3.9 ± 2.7 and distribution of comorbidities as follows: hypertension (84%), heart failure (35%), diabetes (33%), depression (20%), cancer (20%), and anemia (15%). The mean length of stay for the index admission was 4.1 ± 6.1 days and 4.2 ± 6.2 days for the readmission. A small percentage of the patients ever participated in pulmonary rehabilitation before the index admission (6%) or received a palliative care consult during the hospitalization (6%). Approximately 71% of patients reported no moderate to vigorous physical activity (MVPA inactive), 17% were insufficiently active, and another 10% reported engaging in at least 150 minutes of MVPA before their index admission.
| Patient-Level Variables | 30-Day Readmission | Total Index (n = 4,596) | P Value | |
|---|---|---|---|---|
| No (n = 3,653) | Yes (n = 941) | |||
| Age | < 0.01 | |||
| Mean (SD) | 72.1 (10.8) | 73.0 (10.6) | 72.3 (10.8) | |
| Sex | 0.36 | |||
| Male | 1,609 (44) | 431 (46) | 2,040 (44) | |
| Female | 2,044 (56) | 512 (54) | 2,556 (56) | |
| Marital status | < 0.01 | |||
| Unpartnered | 1,735 (47) | 497 (53) | 2,232 (49) | |
| Partnered | 1,749 (48) | 410 (43) | 2,159 (47) | |
| Missing | 169 (5) | 36 (4) | 205 (4) | |
| BMI | 0.67 | |||
| Underweight (<18.5) | 273 (7) | 81 (9) | 354 (8) | |
| Normal weight (18.5–24.9) | 1,166 (32) | 305 (32) | 1,471 (32) | |
| Overweight (25–29.9) | 950 (26) | 242 (26) | 1,192 (26) | |
| Obese (>30) | 1,261 (35) | 315 (33) | 1,576 (34) | |
| Missing | 3 (0) | 0 (0) | 3 (0) | |
| Race | 0.23 | |||
| White | 2,491 (68) | 627 (66) | 3,118 (68) | |
| Black | 513 (14) | 162 (17) | 675 (15) | |
| Hispanic | 438 (12) | 95 (10) | 533 (12) | |
| Asian/Pacific Islander | 154 (4) | 41 (4) | 195 (4) | |
| Others | 19 (1) | 4 (0) | 23 (1) | |
| Missing | 38 (1) | 14 (1) | 52 (1) | |
| Smoking status | 0.001 | |||
| Never | 468 (13) | 100 (11) | 568 (12) | |
| Passive | 22 (1) | 5 (1) | 27 (1) | |
| Former | 2,322 (64) | 666 (71) | 2,988 (65) | |
| Current | 831 (23) | 172 (18) | 1,003 (22) | |
| Missing | 10 (0) | 0 (0) | 10 (0) | |
| Hospitalization in previous year | <0.001 | |||
| No. of patients with no prior hospitalization | 2,228 (61) | 417 (44) | 2,645 (58) | |
| No. of patients with prior COPD hospitalization | 755 (21) | 294 (31) | 1,049 (23) | |
| No. of patients with prior non-COPD hospitalization | 670 (18) | 232 (25) | 902 (20) | |
| FEV1% predicted | 0.13 | |||
| N | 1,976 | 513 | 2,489 | |
| Mean (SD) | 57.7 (21.9) | 56.0 (21.4) | 57.4 (21.8) | |
| FEV1% predicted | 0.70 | |||
| GOLD I | 307 (8) | 69 (7) | 376 (8) | |
| GOLD II | 879 (24) | 235 (25) | 1,114 (24) | |
| GOLD III | 588 (16) | 157 (17) | 745 (16) | |
| GOLD IV | 202 (6) | 52 (6) | 254 (6) | |
| Missing | 1,677 (46) | 430 (46) | 2,107 (46) | |
| Medications | ||||
| Short-acting β-agonist | 3,367 (92) | 870 (92) | 4,237 (92) | 0.93 |
| Short-acting anticholinergic | 2,000 (55) | 584 (62) | 2,584 (56) | <0.001 |
| LABA | 1,538 (42) | 436 (46) | 1,974 (43) | 0.02 |
| LAMA | 1,258 (34) | 377 (40) | 1,635 (36) | < 0.01 |
| ICS | 2,750 (75) | 732 (78) | 3,482 (76) | 0.13 |
| LAMA & ICS | 1,095 (30) | 335 (36) | 1,430 (31) | <0.01 |
| LABA & ICS | 1,527 (42) | 431 (46) | 1,958 (43) | 0.03 |
| Systemic corticosteroids (long-term use) | 368 (10) | 142 (15) | 510 (11) | <0.001 |
| Methylxanthines | 119 (3) | 28 (3) | 147 (3) | 0.65 |
| Oxygen use before index admission | 659 (18) | 188 (20) | 847 (18) | 0.18 |
| New oxygen use post index admission | 414 (11) | 129 (14) | 543 (12) | 0.05 |
| Comorbidities | ||||
| Anemia (present on admission) | 478 (13) | 220 (23) | 698 (15) | <0.001 |
| Heart failure | 1,195 (33) | 436 (46) | 1,631 (35) | <0.001 |
| Pulmonary hypertension | 424 (12) | 150 (16) | 574 (12) | <0.001 |
| Diabetes | 1,156 (32) | 370 (39) | 1,526 (33) | <0.001 |
| Hypertension | 3,028 (83) | 813 (86) | 3,841 (84) | 0.01 |
| Depression | 733 (20) | 207 (22) | 940 (20) | 0.20 |
| Anxiety | 250 (7) | 81 (9) | 331 (7) | 0.06 |
| Solid tumor without metastasis | 659 (18) | 168 (18) | 827 (18) | 0.87 |
| Metastatic cancer | 73 (2) | 28 (3) | 101 (2) | 0.07 |
| Charlson Index | <0.001 | |||
| N | 3,648 | 942 | 4,590 | |
| Mean (SD) | 3.7 (2.62) | 4.5 (2.80) | 3.8 (2.68) | |
| Insurance at index admission | 0.02 | |||
| Medicare | 2,655 (73) | 720 (76) | 3,375 (73) | |
| Medicaid | 59 (2) | 23 (2) | 82 (2) | |
| Commercial | 873 (24) | 187 (20) | 1,060 (23) | |
| Private pay | 30 (1) | 7 (1) | 37 (1) | |
| Missing | 36 (1) | 6 (1) | 42 (1) | |
| Immunization | ||||
| Flu vaccination since Jan. 1, 2010 | 2,828 (77) | 760 (81) | 3,588 (78) | 0.034 |
| Pneumonia vaccination within last 5 yr | 2,311 (63) | 658 (70) | 2,969 (65) | <0.001 |
| Ever participated in pulmonary rehab | 0.93 | |||
| Yes | 212 (6) | 54 (6) | 266 (6) | |
| No | 3,441 (94) | 889 (94) | 4,330 (94) | |
| 30 Day Readmission | Total Index (n = 5,862) | P Value | ||
|---|---|---|---|---|
| No (n = 4,803) | Yes (n = 1,059) | |||
| Length of stay, d | <0.001 | |||
| 1–2 | 1,108 (23) | 193 (18) | 1,301 (22) | |
| 3 | 1,133 (24) | 215 (20) | 1,348 (23) | |
| 4–6 | 1,735 (36) | 396 (37) | 2,131 (36) | |
| 7–13 | 705 (15) | 194 (18) | 899 (15) | |
| ≥14 | 122 (3) | 61 (6) | 183 (3) | |
| Discharge disposition | ||||
| Home | 4,284 (89) | 900 (85) | 5,184 (88) | <0.001 |
| Postacute | 292 (6) | 102 (10) | 394 (7) | |
| Hospital | 7 (0) | 1 (0) | 8 (0) | |
| Hospice | 139 (3) | 9 (1) | 148 (3) | |
| Other | 38 (1) | 20 (2) | 58 (1) | |
| Missing | 43 (1) | 27 (3) | 70 (1) | |
| Emergency dept./observational stay | <0.001 | |||
| Yes | 436 (9) | 198 (19) | 634 (11) | |
| No | 4,367 (91) | 861 (81) | 5,228 (89) | |
| Inpatient palliative care consult | 0.54 | |||
| Yes | 276 (6) | 66 (6) | 342 (6) | |
| No | 4,527 (94) | 993 (94) | 5,520 (94) | |
| Usual physical activity | <0.001 | |||
| 0 min/wk | 3,316 (69) | 827 (78) | 4,143 (71) | |
| 1–149 min/wk | 866 (18) | 141 (13) | 1,007 (17) | |
| >150 min/wk | 497 (10) | 77 (7) | 574 (10) | |
| Missing | 124 (3) | 14 (1) | 138 (2) | |
The all-cause 30-day readmission rate was 18%, with 59% of readmissions occurring in the first 15 days after discharge (Figure 2). To ensure that we excluded any potential transfers that were not already captured, we further excluded any readmission that occurred within 24 hours of the discharge. The distribution across the 10 most common readmission diagnoses (24) during cumulative periods after discharge (Days 0–3, 0–7, 0–15, and 0–30) is shown in Figure 3. Approximately 45% of the readmissions were associated with a principal discharge diagnosis of COPD, followed by other respiratory/ventilator disorders, pneumonia, gastrointestinal disorders, sepsis/shock, heart failure, arrhythmias/conduction disorders, renal disorders, acute myocardial infarction, electrolyte disturbances, and nutrition/metabolic disorders. The overall pattern of readmission diagnoses was largely similar over the cumulative periods after discharge.
Figure 2. Thirty-day readmissions by day (0–30) after hospitalization for chronic obstructive pulmonary disease (COPD).
[More] [Minimize]
Figure 3. Distribution of 30-day readmission principal diagnoses after hospitalizations for chronic obstructive pulmonary disease (COPD). cond = conduction; GI = gastrointestinal; MI = myocardial infarction; resp./vent. = respiratory/ventilatory.
[More] [Minimize]Sex, body mass index, race/ethnicity, use of combination inhalers, use of oxygen supplementation before the index admission, depression, anxiety, and receipt of an inpatient palliative care consult during the index admission were not significantly associated with risk of 30-day readmission (Table 3). Factors associated with a significantly increased 30-day readmission included age, being unpartnered, having a history of any hospitalization in the previous year, new oxygen prescription at discharge, use of systemic steroids, presence of anemia at admission, pulmonary hypertension, hypertension, having more comorbidities, having a visit to the emergency department or an observational stay within 30 days of discharge that occurred before the readmission, and length of stay longer than 3 days (all, P < 0.05). Discharge to postacute care or another hospital was associated with increased risk of readmission (P < 0.001), whereas discharge to hospice was associated with lower risk (P < 0.001). Participation in any MVPA before the index admission was also associated with lower risk of readmission (P < 0.001).
| Parameter | RR | 95% CI | P Value | ||
|---|---|---|---|---|---|
| Age | Age: per increase of 1 yr | 1.01 | 1.00 | 1.01 | 0.03 |
| Sex | Female vs. male | 0.95 | 0.82 | 1.09 | 0.45 |
| Marital status | Unpartnered vs. partnered | 1.20 | 1.04 | 1.38 | 0.01 |
| BMI | Normal weight (18.5–24.9) | 1.00 | |||
| Underweight (<18.5) | 1.10 | 0.84 | 1.44 | 0.48 | |
| Overweight (25–29.9) | 0.97 | 0.80 | 1.16 | 0.71 | |
| Obese (>30) | 0.98 | 0.82 | 1.16 | 0.81 | |
| Race/ethnicity | White | 1.00 | |||
| Asian/Pacific Islander | 0.84 | 0.51 | 1.40 | 0.50 | |
| Black | 0.87 | 0.48 | 1.57 | 0.64 | |
| Hispanic | 0.98 | 0.58 | 1.66 | 0.95 | |
| Others | 0.75 | 0.44 | 1.29 | 0.30 | |
| Insurance type | Commercial/private pay vs. Medicare/Medicaid | 0.81 | 0.69 | 0.96 | 0.02 |
| Smoking status | Never/passive | 1.00 | |||
| Former | 1.27 | 1.02 | 1.59 | 0.03 | |
| Current | 0.96 | 0.74 | 1.25 | 0.79 | |
| Vaccinations | Flu vaccination within 2 yr vs. none | 1.16 | 0.97 | 1.38 | 0.10 |
| Pneumovax within 5 yr vs. none | 1.33 | 1.15 | 1.55 | <0.001 | |
| Usual physical activity | 0 min/wk MVPA (inactive) | 1.00 | |||
| 1-149 min/wk MVPA (insufficient) | 0.67 | 0.55 | 0.81 | <0.0001 | |
| ≥150 min/wk MVPA (active) | 0.61 | 0.47 | 0.80 | <0.0001 | |
| Medications | LABA/ICS or LAMA/ICS vs. none | 1.04 | 0.91 | 1.20 | 0.54 |
| Systemic corticosteroids vs. none | 1.49 | 1.23 | 1.82 | <0.0001 | |
| Oxygen use | Oxygen before index admission vs. none | 1.07 | 0.90 | 1.28 | 0.42 |
| New oxygen post index admission vs. none | 1.29 | 1.05 | 1.58 | 0.02 | |
| Hospitalizations | No previous hospitalizations | 1.00 | |||
| Prior COPD hospitalization | 1.82 | 1.55 | 2.13 | <0.0001 | |
| Prior non-COPD hospitalization | 1.82 | 1.53 | 2.17 | <0.0001 | |
| Comorbidities | Charlson index score, 0–1 | 1.00 | |||
| Charlson index score, 2 | 1.14 | 0.88 | 1.46 | 0.32 | |
| Charlson index score, 3+ | 1.88 | 1.55 | 2.28 | <0.0001 | |
| Pulmonary hypertension vs. none | 1.39 | 1.14 | 1.68 | <0.0001 | |
| Hypertension vs. none | 1.32 | 1.09 | 1.61 | 0.01 | |
| Depression vs. none | 1.12 | 0.95 | 1.33 | 0.18 | |
| Anxiety vs. none | 1.22 | 0.96 | 1.56 | 0.11 | |
| Anemia (present on admission) vs. none | 1.94 | 1.63 | 2.29 | <0.0001 | |
| Chronic obstructive asthma vs. COPD | 0.95 | 0.82 | 1.11 | 0.55 | |
| Palliative care | Received consult vs. no | 1.06 | 0.80 | 1.42 | 0.69 |
| Length of stay, d | 1–2 | 1.00 | |||
| 3 | 1.09 | 0.88 | 1.35 | 0.44 | |
| 4–6 | 1.29 | 1.07 | 1.56 | 0.01 | |
| 7-13 | 1.52 | 1.21 | 1.91 | 0.00 | |
| ≥14 | 2.79 | 1.98 | 3.94 | <0.0001 | |
| ED/observational stay | Yes vs. no | 2.23 | 1.85 | 2.69 | <0.0001 |
| Disposition | Home | 1.00 | |||
| Hospice | 0.28 | 0.13 | 0.04 | <0.0001 | |
| Postacute/hospital | 1.65 | 1.30 | 61.56 | <0.0001 | |
| Other/missing | 2.76 | 1.90 | 210.61 | <0.0001 | |
After adjustments for all other key covariates, patients who reported any MVPA at baseline before the index admission had a significantly lower risk of 30-day readmission compared with inactive patients (Table 4). Specifically, patients meeting the national recommendation of engaging in at least 150 min/wk of MVPA had a 34% lower risk of 30-day readmission compared with inactive patients (relative risk [RR], 0.66; 95% confidence interval [CI], 0.51–0.87). Patients engaged in lower levels of MVPA were also less likely to be readmitted (RR, 0.67; 95% CI, 0.55–0.81). Factors that remained significantly associated with increased risk of 30-day readmission in this adjusted model included having a history of any prior hospitalizations, a new prescription for oxygen, presence of anemia at admission, more comorbidities, a visit to the emergency department or an observational stay, and length of stay longer than 3 days (all, P < 0.01); being unpartnered approached statistical significance (P = 0.08).
| Parameter | RR | 95% CI | P Value | ||
|---|---|---|---|---|---|
| Age | Age: per increase of 1 yr | 1.00 | 0.99 | 1.01 | 0.89 |
| Sex | Female vs. male | 0.95 | 0.81 | 1.11 | 0.48 |
| Marital status | Unpartnered vs. partnered | 1.15 | 0.98 | 1.34 | 0.08 |
| BMI | Normal weight (18.5–24.9) | 1.00 | |||
| Underweight (<18.5) | 1.08 | 0.82 | 1.42 | 0.57 | |
| Overweight (25–29.9) | 0.92 | 0.76 | 1.11 | 0.40 | |
| Obese (>30) | 0.87 | 0.72 | 1.05 | 0.14 | |
| Race/ethnicity | White | 1.00 | |||
| Asian/Pacific Islander | 0.80 | 0.46 | 1.37 | 0.41 | |
| Black | 0.92 | 0.48 | 1.74 | 0.79 | |
| Hispanic | 0.85 | 0.48 | 1.51 | 0.59 | |
| Others | 0.71 | 0.40 | 1.28 | 0.26 | |
| Insurance type | Commercial/private pay vs. Medicare/Medicaid | 0.95 | 0.77 | 1.17 | 0.64 |
| Smoking status | Never/passive | 1.00 | |||
| Former | 1.25 | 0.99 | 1.57 | 0.06 | |
| Current | 1.06 | 0.80 | 1.40 | 0.67 | |
| Vaccinations | Flu vaccination within 2 yr vs. none | 0.95 | 0.79 | 1.14 | 0.59 |
| Pneumovax within 5 yr vs. none | 1.20 | 1.02 | 1.41 | 0.03 | |
| Usual physical activity | 0 min/wk MVPA(inactive) | 1.00 | |||
| 1–149 min/wk MVPA (insufficient) | 0.67 | 0.55 | 0.81 | <0.0001 | |
| ≥150 min/wk MVPA (active) | 0.66 | 0.51 | 0.87 | <0.0001 | |
| Medications | LABA/ICS or LAMA/ICS vs. none | 0.94 | 0.81 | 1.09 | 0.39 |
| Systemic corticosteroids vs. none | 1.24 | 1.01 | 1.52 | 0.04 | |
| Oxygen use | Oxygen before index admission vs. none | 0.93 | 0.77 | 1.12 | 0.43 |
| New oxygen post index admission vs. none | 1.52 | 1.22 | 1.90 | <0.0001 | |
| Hospitalizations | No previous hospitalizations | 1.00 | |||
| Prior COPD hospitalization | 1.53 | 1.29 | 1.82 | <0.0001 | |
| Prior non-COPD hospitalization | 1.60 | 1.31 | 1.94 | <0.0001 | |
| Comorbidities | Charlson index score, 0–1 | 1.00 | |||
| Charlson index score, 2 | 0.98 | 0.76 | 1.26 | 0.88 | |
| Charlson index score, 3+ | 1.44 | 1.17 | 1.78 | <0.0001 | |
| Pulmonary hypertension vs. none | 1.06 | 0.86 | 1.31 | 0.57 | |
| Hypertension vs. none | 0.94 | 0.76 | 1.16 | 0.55 | |
| Depression vs. none | 0.95 | 0.79 | 1.13 | 0.55 | |
| Anxiety vs. none | 1.13 | 0.89 | 1.45 | 0.32 | |
| Anemia (present on admission) vs. None | 1.59 | 1.33 | 1.90 | <0.0001 | |
| Chronic obstructive asthma vs. COPD | 1.03 | 0.87 | 1.23 | 0.72 | |
| Palliative care | Received consult vs. No | 0.86 | 0.64 | 1.17 | 0.34 |
| Length of stay, d | 1–2 | 1.00 | |||
| 3 | 1.17 | 0.94 | 1.46 | 0.15 | |
| 4–6 | 1.37 | 1.13 | 1.66 | <0.0001 | |
| 7–13 | 1.66 | 1.30 | 2.10 | <0.0001 | |
| ≥14 | 2.93 | 1.99 | 4.31 | <0.0001 | |
| ED/observational stay | Yes vs. no | 2.13 | 1.76 | 2.59 | <0.0001 |
| Disposition | Home | 1.00 | |||
| Hospice | 0.25 | 0.12 | 0.52 | <0.0001 | |
| Postacute/hospital | 1.24 | 0.96 | 1.61 | 0.10 | |
| Other/missing | 3.13 | 2.05 | 4.79 | <0.0001 | |
We performed sensitivity analyses to determine if the effects of regular physical activity on 30-day readmission would remain robust to three different scenarios. First, we analyzed either the median or modal value of all EVS data in the 12 months preceding the hospitalization and found similar effects (see online supplement). Second, we excluded patients with a principal diagnosis of acute respiratory failure and a secondary diagnosis of COPD. The effect of physical inactivity was similar in this analysis (insufficiently active vs. inactive: RR, 0.67; 95% CI, 0.54–0.82; active vs. inactive: RR, 0.65; 95% CI, 0.49–0.86). Third, we included FEV1% predicted as a continuous variable and as Global Initiative for Chronic Obstructive Lung Disease stages in the model to further adjust for disease severity in the 54% of the analytical sample with spirometry data; after adjustment for FEV1% predicted, the effect of regular physical activity was only slightly attenuated (insufficiently active vs. inactive: RR, 0.68; 95% CI, 0.52–0.88; active vs. inactive: RR, 0.79; 95% CI, 0.56–1.1). Patients who had spirometry data were significantly younger as well as more likely to be men and on supplemental oxygen compared with those without spirometry data.
A significant novel finding from our study was that patients admitted for an acute COPD exacerbation who had been engaging in any level of MVPA before their hospitalization had a 34% lower risk of being readmitted within 30 days of the index admission compared with inactive patients. This study is especially ground-breaking because our measure of regular physical activity was derived from routine clinical care in contrast to earlier studies that either administered lengthy physical activity questionnaires or used more cumbersome objective measurements in smaller research samples (14–19). An important implication of our finding is that current efforts to prevent 30-day readmission should be broadened beyond treating the acute COPD exacerbation and its complications or addressing failures with the postdischarge transition process. A shift to more proactive management of modifiable lifestyle behaviors to both decrease the risk of having the initial hospitalization and optimize patients’ chances of a successful recovery should they become hospitalized is needed.
Physical inactivity is highly prevalent in patients with stable COPD (25, 26). Population-based data using accelerometry from the National Health and Nutrition Survey showed that patients with COPD accrued on average only 6 minutes of MVPA per day, which is far below the national recommendation for adults (25). Patients are markedly more inactive during and after hospitalizations for COPD, and those having persistently low activity 1 month after discharge are at significantly greater risk of being readmitted within the following year (16, 19, 27). These observations beg the question of whether or not physical inactivity is an independent causal determinant of worse outcomes and could be targeted for intervention or if it is merely another marker of disease severity (28). Findings from a recent Cochrane systematic review showing that pulmonary rehabilitation provided during or soon after a COPD exacerbation was associated with a lower risk of future hospitalizations (though not specifically 30-d readmission) suggest that increasing physical activity is likely to reduce the risk of hospitalizations (29).
Our findings resonate with Krumholz’s (30) premise that patients likely experience “post-hospital syndrome, an acquired transient period of vulnerability” and are at generalized risk for a myriad of conditions due to the allostatic and physiological stress associated with the hospitalization on top of the effects of their acute illness. Yet, it is possible that the stresses of hospitalization could be buffered by an overall, more physically conditioned state. Although admissions for COPD exacerbations for the most part are not planned, studies of elective surgical patients have demonstrated the importance of prehabilitating patients before major surgical procedures to optimize recovery (31–33) (e.g., pulmonary rehabilitation is required before and after lung volume reduction surgery [34] or lung transplant). Interestingly, we did not find that previous participation in pulmonary rehabilitation was associated with a lower risk of 30-day readmission. The lack of observed effects could be due to the relatively small number of patients who participated in rehabilitation and the inclusion of patients who had ever participated in rehabilitation in the 4 years preceding their index admission (where the effects would not be expected to persist). More importantly, what our data suggest is that, in the long run, reducing risk of readmission is likely less dependent on participation in pulmonary rehabilitation but more influenced by maintenance of a physically active lifestyle in the face of a progressive illness (35).
Other significant independent predictors of readmission, such as having a history of prior hospitalizations (10, 36), presence of anemia at admission (37, 38), having more comorbidities (10, 37), and longer length of stay (10), are consistent with previous studies that examined risk of hospitalizations in COPD for periods longer than 30 days after discharge. Receiving a new oxygen prescription and having a visit to the emergency department or observational stay before the readmission also increased risk but are potentially modifiable. Patients who received supplemental oxygen at discharge may have been more physiologically compromised, were previously undertreated for their hypoxemia (36), or may have had challenges to or did not adhere to using their oxygen as prescribed. If it is due to the latter two reasons, there are opportunities to intervene through more intensive outpatient management and/or patient education. The twofold risk of readmission for patients who used the emergency department or were cared for in an observational stay unit in the days preceding the readmission is noteworthy and possibly reflects a combination of inadequate treatment for the presenting condition or an unintended consequence of existing performance-based reimbursement policies (39, 40). Further study is needed to better understand the underlying drivers. Finally, being unpartnered was associated with a trend toward greater risk for readmission, which likely reflects inadequate social support for engagement in self-care (41).
Strengths of our study include a large, contemporary, diverse, and sex-balanced sample, as well as the innovative use of physical activity data that are routinely captured during every outpatient visit, complete capture of all health resource use given our comprehensive claims database, and estimates of the effects of physical activity on 30-day readmission that remain robust to several sensitivity analyses.
There are several limitations to this study. First, we relied on administrative and clinical records, specifically, use of the EVS, which is routinely collected by front-line clinical staff during all outpatient encounters. It is possible that patients who were ultimately admitted for treatment of their COPD exacerbation were seen in the outpatient setting in the weeks before their admission and that the EVS data reflected the diminished physical activity associated with their current exacerbation. However, findings from our sensitivity analysis where all EVS data from the 12 months before the hospitalization were used further confirm that the association between the EVS and 30-day readmission reflects patients’ regular levels of physical activity, not acute impairments in their activity levels associated with an exacerbation. Second, as described earlier, physical inactivity could merely be another marker of disease severity. Third, because we relied on routinely collected administrative and clinical data, we were not able to fully adjust for a number of other covariates that were either incompletely documented or not easily accessible (e.g., airflow obstruction, exacerbation severity [14], in-hospital complications, or socioeconomic factors [42]), all of which have been shown to be associated with higher readmission risk. Finally, the cohort was drawn from an insured population within a financially aligned and integrated health system, and thus the findings may not be generalizable to patients with COPD who are uninsured or have fee-for-service public insurance.
A key novel finding from our study was that self-reported participation in any level of regular moderate to vigorous physical activity captured as part of routine clinical care was associated with a significantly lower risk of all-cause 30-day readmission in a large cohort of patients with COPD. Although it is possible that physical inactivity is a reflection of worse disease, promoting and supporting physical activity is a promising strategy not only to reduce the risk of having the index hospitalization but also to buffer the stresses of hospitalization. Our findings further support the importance of physical activity in the overall management of COPD across the care continuum, including care transition efforts to reduce 30-day readmissions.
| 1 . | Centers for Medicare and Medicaid Services. Hospital readmissions reduction program. [modified 2013 Aug 2; accessed 2013 Dec 28]. Available from: http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/Readmissions-Reduction-Program.html |
| 2 . | Kocher RP, Adashi EY. Hospital readmissions and the Affordable Care Act: paying for coordinated quality care. JAMA 2011;306:1794–1795. |
| 3 . | Bradley EH, Sipsma H, Horwitz LI, Curry L, Krumholz HM. Contemporary data about hospital strategies to reduce unplanned readmissions: what has changed? JAMA Intern Med 2014;174:154–156. |
| 4 . | Brock J, Mitchell J, Irby K, Stevens B, Archibald T, Goroski A, Lynn J; Care Transitions Project Team. Association between quality improvement for care transitions in communities and rehospitalizations among Medicare beneficiaries. JAMA 2013;309:381–391. |
| 5 . | Hansen LO, Young RS, Hinami K, Leung A, Williams MV. Interventions to reduce 30-day rehospitalization: a systematic review. Ann Intern Med 2011;155:520–528. |
| 6 . | Rodriguez-Roisin R. Toward a consensus definition for COPD exacerbations. Chest 2000; 117:398S–401S. |
| 7 . | Dalal AA, Christensen L, Liu F, Riedel AA. Direct costs of chronic obstructive pulmonary disease among managed care patients. Int J Chron Obstruct Pulmon Dis 2010;5:341–349. |
| 8 . | Almagro P, Calbo E, Ochoa de Echagüen A, Barreiro B, Quintana S, Heredia JL, Garau J. Mortality after hospitalization for COPD. Chest 2002;121:1441–1448. |
| 9 . | Connors AF Jr, Dawson NV, Thomas C, Harrell FE Jr, Desbiens N, Fulkerson WJ, Kussin P, Bellamy P, Goldman L, Knaus WA. Outcomes following acute exacerbation of severe chronic obstructive lung disease. The SUPPORT investigators (Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments). Am J Respir Crit Care Med 1996;154:959–967. |
| 10 . | McGhan R, Radcliff T, Fish R, Sutherland ER, Welsh C, Make B. Predictors of rehospitalization and death after a severe exacerbation of COPD. Chest 2007;132:1748–1755. |
| 11 . | Elixhauser A, Au DH, Podulka J. Readmissions for chronic obstructive pulmonary disease, 2008: Statistical Brief #121. Healthcare Cost and Utilization Project (HCUP) Statistical Briefs. Rockville, MD: Agency for Health Care Policy and Research (US); 2011. |
| 12 . | Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med 2009;360:1418–1428. |
| 13 . | Kansagara D, Englander H, Salanitro A, Kagen D, Theobald C, Freeman M, Kripalani S. Risk prediction models for hospital readmission: a systematic review. JAMA 2011;306:1688–1698. |
| 14 . | Garcia-Aymerich J, Farrero E, Félez MA, Izquierdo J, Marrades RM, Antó JM; Estudi del Factors de Risc d’Agudització de la MPOC investigators. Risk factors of readmission to hospital for a COPD exacerbation: a prospective study. Thorax 2003;58:100–105. |
| 15 . | Garcia-Aymerich J, Lange P, Benet M, Schnohr P, Antó JM. Regular physical activity reduces hospital admission and mortality in chronic obstructive pulmonary disease: a population based cohort study. Thorax 2006;61:772–778. |
| 16 . | Pitta F, Troosters T, Probst VS, Spruit MA, Decramer M, Gosselink R. Physical activity and hospitalization for exacerbation of COPD. Chest 2006;129:536–544. |
| 17 . | Waschki B, Kirsten A, Holz O, Müller KC, Meyer T, Watz H, Magnussen H. Physical activity is the strongest predictor of all-cause mortality in patients with COPD: a prospective cohort study. Chest 2011;140:331–342. |
| 18 . | Moy ML, Teylan M, Weston NA, Gagnon DR, Garshick E. Daily step count predicts acute exacerbations in a US cohort with COPD. PLoS ONE 2013;8:e60400. |
| 19 . | Zanoria SJ, ZuWallack R. Directly measured physical activity as a predictor of hospitalizations in patients with chronic obstructive pulmonary disease. Chron Respir Dis 2013;10:207–213. |
| 20 . | Nguyen HQ, Chu L, Liu IL, Lee J, Suh D, Korotzer B, Yuen G, Gould MK. Higher level of regular physical activity is associated with lower risk of 30-day readmissions in patients with COPD [abstract]. Presented at the American Thoracic Society 2014 International Conference. 2014; San Diego, CA. |
| 21 . | National Quality Forum. NQF #1891 Hospital 30-day, all-cause, risk-standardized readmission rate (RSRR) following chronic obstructive pulmonary disease (COPD) hospitalization. 2012 [accessed 2013 Dec 28]. http://www.qualityforum.org/WorkArea/linkit.aspx?LinkIdentifier=id&ItemID=69926 |
| 22 . | Coleman KJ, Ngor E, Reynolds K, Quinn VP, Koebnick C, Young DR, Sternfeld B, Sallis RE. Initial validation of an exercise “vital sign” in electronic medical records. Med Sci Sports Exerc 2012;44:2071–2076. |
| 23 . | Grant RW, Schmittdiel JA, Neugebauer RS, Uratsu CS, Sternfeld B. Exercise as a vital sign: a quasi-experimental analysis of a health system intervention to collect patient-reported exercise levels. J Gen Intern Med 2014;29:341–348. |
| 24 . | Dharmarajan K, Hsieh AF, Lin Z, Bueno H, Ross JS, Horwitz LI, Barreto-Filho JA, Kim N, Bernheim SM, Suter LG, et al. Diagnoses and timing of 30-day readmissions after hospitalization for heart failure, acute myocardial infarction, or pneumonia. JAMA 2013;309:355–363. |
| 25 . | Park SK, Richardson CR, Holleman RG, Larson JL. Physical activity in people with COPD, using the National Health and Nutrition Evaluation Survey dataset (2003-2006). Heart Lung 2013;42:235–240. |
| 26 . | Watz H, Waschki B, Meyer T, Magnussen H. Physical activity in patients with COPD. Eur Respir J 2009;33:262–272. |
| 27 . | Ehsan M, Khan R, Wakefield D, Qureshi A, Murray L, Zuwallack R, Leidy NK. A longitudinal study evaluating the effect of exacerbations on physical activity in patients with chronic obstructive pulmonary disease. Ann Am Thorac Soc 2013;10:559–564. |
| 28 . | Polkey MI, Rabe KF. Chicken or egg: physical activity in COPD revisited. Eur Respir J 2009;33:227–229. |
| 29 . | Puhan MA, Gimeno-Santos E, Scharplatz M, Troosters T, Walters EH, Steurer J. Pulmonary rehabilitation following exacerbations of chronic obstructive pulmonary disease. Cochrane Database Syst Rev 2011;10:CD005305. |
| 30 . | Krumholz HM. Post-hospital syndrome—an acquired, transient condition of generalized risk. N Engl J Med 2013;368:100–102. |
| 31 . | Li C, Carli F, Lee L, Charlebois P, Stein B, Liberman AS, Kaneva P, Augustin B, Wongyingsinn M, Gamsa A, et al. Impact of a trimodal prehabilitation program on functional recovery after colorectal cancer surgery: a pilot study. Surg Endosc 2013;27:1072–1082. |
| 32 . | Carli F, Charlebois P, Stein B, Feldman L, Zavorsky G, Kim DJ, Scott S, Mayo NE. Randomized clinical trial of prehabilitation in colorectal surgery. Br J Surg 2010;97:1187–1197. |
| 33 . | Silver JK, Baima J. Cancer prehabilitation: an opportunity to decrease treatment-related morbidity, increase cancer treatment options, and improve physical and psychological health outcomes. Am J Phys Med Rehabil 2013;92:715–727. |
| 34 . | Fishman A, Martinez F, Naunheim K, Piantadosi S, Wise R, Ries A, Weinmann G, Wood DE; National Emphysema Treatment Trial Research Group. A randomized trial comparing lung-volume-reduction surgery with medical therapy for severe emphysema. N Engl J Med 2003;348:2059–2073. |
| 35 . | Pitta F, Troosters T, Probst VS, Langer D, Decramer M, Gosselink R. Are patients with COPD more active after pulmonary rehabilitation? Chest 2008;134:273–280. |
| 36 . | Garcia-Aymerich J, Monsó E, Marrades RM, Escarrabill J, Félez MA, Sunyer J, Antó JM; EFRAM Investigators. Risk factors for hospitalization for a chronic obstructive pulmonary disease exacerbation. EFRAM study. Am J Respir Crit Care Med 2001;164:1002–1007. |
| 37 . | Almagro P, Cabrera FJ, Diez J, Boixeda R, Alonso Ortiz MB, Murio C, Soriano JB; Working Group on COPD, Spanish Society of Internal Medicine. Comorbidities and short-term prognosis in patients hospitalized for acute exacerbation of COPD: the EPOC en Servicios de medicina interna (ESMI) study. Chest 2012;142:1126–1133. |
| 38 . | Barba R, de Casasola GG, Marco J, Emilio Losa J, Plaza S, Canora J, Zapatero A. Anemia in chronic obstructive pulmonary disease: a readmission prognosis factor. Curr Med Res Opin 2012;28:617–622. |
| 39 . | Feng Z, Wright B, Mor V. Sharp rise in Medicare enrollees being held in hospitals for observation raises concerns about causes and consequences. Health Aff (Millwood) 2012;31:1251–1259. |
| 40 . | Sheehy AM, Graf B, Gangireddy S, Hoffman R, Ehlenbach M, Heidke C, Fields S, Liegel B, Jacobs EA. Hospitalized but not admitted: characteristics of patients with “observation status” at an academic medical center. JAMA Intern Med 2013;173:1991–1998. |
| 41 . | Bucknall CE, Miller G, Lloyd SM, Cleland J, McCluskey S, Cotton M, Stevenson RD, Cotton P, McConnachie A. Glasgow supported self-management trial (GSuST) for patients with moderate to severe COPD: randomised controlled trial. BMJ 2012;344:e1060. |
| 42 . | Calvillo-King L, Arnold D, Eubank KJ, Lo M, Yunyongying P, Stieglitz H, Halm EA. Impact of social factors on risk of readmission or mortality in pneumonia and heart failure: systematic review. J Gen Intern Med 2013;28:269–282. |
Supported by the Kaiser Permanente Southern California, Implementation Science Program.
Author Contributions: H.Q.N. contributed substantially to the study design, analysis, interpretation, and the preparation of this manuscript. H.Q.N. had full access to the data and will vouch for the integrity of the work as a whole, from inception to published article. L.C. contributed to the data acquisition, interpretation, and the preparation of this manuscript. I.-L.A.L. contributed to the data analysis, interpretation, and the preparation of this manuscript. J.S.L. contributed to the data acquisition, interpretation, and the preparation of this manuscript. D.S. contributed to the interpretation and the preparation of this manuscript. B.K. contributed to the interpretation and the preparation of this manuscript. G.Y. contributed to the interpretation and the preparation of this manuscript. S.D. contributed to the interpretation and the preparation of this manuscript. K.J.C. contributed to the interpretation and the preparation of this manuscript. A.H.X. contributed to the interpretation and the preparation of this manuscript. M.K.G. contributed to the study design, analysis, interpretation, and the preparation of this manuscript.
This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org
A portion of these findings were submitted in abstract form to the American Thoracic Society 2014 International Conference in San Diego, CA.
Author disclosures are available with the text of this article at www.atsjournals.org.
