Abstract
Objective
To identify individual-level factors associated with hospital readmission among individuals with SSc-associated pulmonary hypertension (SSc-PH).
Methods
Individuals enrolled in the Pulmonary Hypertension Assessment and Recognition of Outcomes in Scleroderma (PHAROS) registry contributed clinical data related to SSc-PH disease severity and hospital admissions. Readmission was defined as a subsequent hospitalization within 12 months of any hospital discharge. Characteristics were compared between individuals with and without readmissions using Fisher’s exact test, Wilcoxon rank-sum test, or Kruskal–Wallis test. Logistic regression was used to estimate associations between clinical predictors and likelihood of readmission.
Results
Of 572 individuals with SSc-PH enrolled in PHAROS, 54% had ≥1 hospitalizations between 2005 and 2016. Among individuals ever-hospitalized, 34% had ≥1 readmission. Individuals with vs without readmissions had shorter median (IQR) time between index hospitalization date and next PHAROS visit [37 (3, 80) vs 81 (42, 136) days, P <0.001]. Index admissions related to PH or SSc (vs non-PH/SSc related) were associated with an increased odds of 12-month readmission [aOR 6.6 (95% CI 3.2, 13.6) and aOR 2.2 (95% CI 1.1, 4.5), respectively]. Readmission was less likely among home oxygen users (vs non-users) (aOR 0.44; 95% CI 0.22, 0.89). Race, age, sex, disease duration and disease subtype were not associated with readmission.
Conclusion
The strongest predictor for 12-month readmission was an index hospitalization reason related to PH. Home oxygen use was associated with lower odds of readmission. Future studies should determine whether testing for the need for home oxygen mediates the risk of readmission in SSc-PH.
Keywords: SSc, pulmonary hypertension, hospitalization, readmission, morbidity, scleroderma
Rheumatology key messages.
Among ever-hospitalized individuals enrolled in PHAROS, 34% had ≥1 hospital readmission within 12 months.
Home oxygen use was associated with lower odds of readmission for those with systemic sclerosis associated pulmonary hypertension within 12 months.
The strongest predictor for 12-month readmission was an index hospitalization reason related to pulmonary hypertension.
Introduction
SSc is an autoimmune condition characterized by inflammation, fibrosis and vasculopathy, and pulmonary hypertension (PH) is among the leading causes of death [1, 2]. The World Health Organization (WHO) classifies PH by pulmonary pressures and aetiology. WHO group 1 refers to pulmonary arterial hypertension (PAH); WHO group 2 refers to pulmonary venous hypertension (e.g. due to left heart failure); WHO group 3 refers to pulmonary hypertension associated with disorders of the respiratory system or hypoxaemia (e.g. due to interstitial lung disease); WHO group 4 refers to pulmonary hypertension caused by chronic thrombotic or embolic disease, and WHO group 5 includes miscellaneous causes (such as sarcoidosis) [3]. SSc is a highly recognized cause of WHO group 1 PH; however, individuals with SSc may have other PH types including group 2 and group 3 [4].
Hospital admissions and readmissions occur frequently in individuals with PAH [5, 6] and are associated with poor outcomes [7]. Using a United States claims database, 2275 hospitalized individuals with PAH were identified [6]. Readmission over the five-year study period was observed in 53% of participants (21% within 30 days and 79% within one year). An observational study of 862 individuals with PAH found that ∼75% of those hospitalized for PAH were readmitted during the three-year study period [7]. Three-year survival was lower for those with either PAH-related or PAH-unrelated hospitalizations [57% (3.5%) and 68% (3.6%), respectively] compared with those who were not hospitalized [78% (1.9%)] [7].
High hospitalization rates are observed among individuals with SSc and occur most often during the initial five years of disease [8]. In Italy, among 4981 SSc hospitalizations between 2001 and 2012, 6.4% were related to PAH [9]. Using the United States Healthcare Cost and Utilization Project National Readmission Database between 2010 and 2015, 22 420 hospitalizations for individuals with SSc-associated PH were identified [5]. Of those, 22% were associated with a 30-day readmission. Independent risk factors for 30-day readmission included male sex, congestive heart failure (CHF), Medicare or Medicaid coverage, and discharge to a skilled nursing facility or short-term hospital (vs to home). However, the study data source lacked details related to SSc and PH severity that may also contribute to readmission risk.
In addition to prognostic implications, hospital readmissions related to PH are associated with increased cost. Among commercially insured patients with PH-related hospitalizations (of any WHO Group), the mean readmission cost was higher than the cost of the initial hospitalization ($73 066 vs $46 070) [6]. Further, 30-day readmissions have been shown to be more costly among those with SSc-PH compared with non-SSc PAH (P <0.001), despite similar charges associated with index hospitalizations [5].
The objective of this retrospective analysis is to identify individual-level factors associated with hospital readmission among well-characterized individuals with SSc enrolled in the Pulmonary Hypertension Assessment and Recognition of Outcomes in Scleroderma (PHAROS) registry.
Participants and methods
Analytic sample
This retrospective cohort study uses clinical data from the prospective, observational PHAROS registry, a multicentre, cohort study designed to understand the natural history of SSc-PH and risk-factors for incident SSc-PH [4]. All participants met ACR classification criteria for SSc or the LeRoy SSc definitions [10, 11]. Though pulmonary pressure diagnostic thresholds for PH have evolved over time, at the time of initial study enrolment (2005), PH was defined as mPAP ≥25mm Hg on right heart catheterization. In 2018, this threshold was reduced to mPAP ≥20mm Hg, but required pulmonary vascular resistance of >3 Woods units [12]. Enrolment in PHAROS was completed by that time. Because one objective of the PHAROS registry was to determine risk factors for incident SSc-PH, individuals without PH who were at high risk for development of PH were included. Those classified as ‘at-risk’ for PH had (i) %-predicted carbon monoxide diffusing capacity (DLCO) <55, (ii) %-predicted forced vital capacity (FVC)/DLCO ratio ≥1.6, or (iii) systolic pulmonary arterial pressure >40 mmHg on echocardiography [13]. This study complies with the Declaration of Helsinki and was approved by each of the participating centre’s local institutional review boards/ethics committees (Supplementary Data Section S1, available at Rheumatology online). All study participants provided written informed consent. Data are available on request.
Clinical variables
Clinical data were analysed from the closest study encounter following the index (initial) hospitalization date except when the post-hospitalization PHAROS visit was significantly farther from the admission than the pre-hospitalization visit (>30 days), in which case pre-hospitalization PHAROS visit data were collected. Variables collected included: age, sex, race, SSc autoantibody profile, SSc disease subtype and duration, WHO PH Group, modified Rodnan skin score (MRSS), history of digital ulcers, pulmonary function test results, home oxygen use and need for oxygen during 6-min walk test. Patient-reported outcomes were collected including the Borg Dyspnoea Index, 36-Item Short Form Survey (SF-36) physical and mental health score [14], University of California at San Diego Dyspnoea Index [15] and Scleroderma HAQ Disability Index [16]. Data were also collected regarding use of the following medications: mycophenolate mofetil, cyclophosphamide, calcium channel blocking agents, prostacyclin analogs, phosphodiesterase type 5 inhibitors and endothelin receptor antagonists.
Hospitalizations
The PHAROS investigators recorded all hospitalizations’ primary admission reason and date. Readmission was defined a priori as a subsequent hospitalization for any reason within one year, because we felt this cut-point to be clinically relevant and prior work in PH has shown survival differences in those with vs without hospitalizations over a study period of three years [7]. A sensitivity analysis was performed to assess factors associated with 30- and 90-day readmissions.
Statistical analysis
Descriptive statistics were performed and percentages, means (standard deviations) and medians (interquartile ranges) are reported. Bivariate comparisons were performed by using Fisher’s exact or χ2 tests for categorical variables and Wilcoxon rank sum or Kruskal–Wallis tests for continuous variables. Logistic regression was performed using both univariable and multivariable models. The multivariable model adjusted for age, sex, race, WHO PH Group, index hospitalization admission reason and home oxygen use. Age, sex and race were selected a priori for inclusion in the model, while the other covariates were selected as they were found to be statistically significant in the bivariate analysis. Multiple imputation with chained equations was performed to account for missing data. Specifically, 26 imputed datasets were created with the seed of 1979 using the mi impute chained function in Stata. Logistic regression results reported are from the multiply imputed dataset [17, 18]. Adjusted odds ratios (aOR) and 95% confidence intervals (95% CI) were calculated for all estimates. An alpha of 0.05 was used to determine statistical significance. All analyses were performed using Stata version 14.
Results
Readmissions
Of 572 individuals included in the PHAROS registry, 307 (54%) had at least one hospitalization over a median (IQR) follow-up (index hospitalization to last PHAROS visit) of 2.6 (1.3, 4.4) years between May 2005 and July 2016. The median (interquartile range) time between index hospitalization and next PHAROS visit was 64 (28, 123) days. For 20 individuals, data were collected from the PHAROS visit before the index hospitalization. Among these 20 individuals, the median (IQR) time from PHAROS visit to index hospitalization was 65 days (20, 82) and time from index hospital admission to next PHAROS visit was 134 days (105, 169). Readmission within 12 months was observed in 104 individuals (34%): 93 had one readmission within 12 months, and 11 had two or more readmissions within 12 months. Readmission within 90 days was observed in 31 individuals, and readmission within 30 days was observed in 17 individuals (Fig. 1).
Fig. 1 .
Flowchart of individuals enrolled in the PHAROS registry according to hospital admissions and readmissions (2005–2016)
Mortality
Among the 307 individuals in the cohort, there were 109 deaths (36%). Among the 109 deaths, 71 (65%) had no readmissions and 38 (35%) had at least one readmission. Specifically, 35 individuals had one readmission, one individual had two readmissions, and two individuals had three readmissions. There was no statistical difference in mortality between those with vs without at least one 12-month readmission (39% vs 35%, P = 0.80). For both those with and without readmission, the most frequently reported cause of death was pulmonary arterial hypertension (37% and 35%, respectively).
Clinical characteristics
Clinical characteristics are presented for individuals with and without readmission in Table 1 and Supplementary Table S1, available at Rheumatology online. Among individuals with (vs without) readmission, median (IQR) time between index hospitalization date and next PHAROS visit was shorter [37 (3, 80) vs 81 (42, 136) days, P <0.001] and median (IQR) time between index hospitalization and last PHAROS visit date was longer [2.8 (1.7, 4.9) vs 2.4 (1.0, 4.1) years, P =0.027]. Fewer individuals with (vs without) readmission used home oxygen (18% vs 30%, P =0.04). There were no statistically significant differences among those with vs without readmission in age, sex, race, scleroderma subtype (limited or diffuse) or pulmonary function test results. Patient-reported outcomes, including the Borg Dyspnoea Index, SF-36 physical and mental health score, University of California at San Diego Dyspnoea Index and Scleroderma HAQ Disability Index were not significantly different between those with vs without readmission (Supplementary Table S1, available at Rheumatology online). Ever-use of cyclophosphamide at the time of first post-hospitalization PHAROS visit was more common among individuals without vs with readmission (10% vs 2%, P =0.015). There were no differences in use of mycophenolate mofetil, calcium channel blocking agents, prostacyclin analogs, phosphodiesterase type 5 inhibitors and endothelin receptor antagonists (Supplementary Table S1, available at Rheumatology online).
Table 1.
Characteristics of individuals enrolled in the PHAROS registry with at least one hospitalization
| Clinical Characteristic | Total (n = 307) | No readmission within 12 months (n = 203) | ≥1 readmission within 12 months (n = 104) | P-value |
|---|---|---|---|---|
| Age, mean (s.d.), years | 57.3 (11.4) | 57.0 (11.6) | 57.8 (11.0) | 0.58 |
| Female, n (%)+ | 248 (80.8%) | 162 (79.8%) | 86 (82.7%) | 1.00 |
| Disease duration, years, median (IQR) | 5.6 (1.6, 11.4) | 5.6 (1.7, 11.3) | 5.6 (1.4, 11.5) | 0.92 |
| Race, n (%) | ||||
| White | 221 (72.0%) | 146 (71.9%) | 75 (72.1%) | 0.82 |
| Black | 53 (17.3%) | 33 (16.3%) | 20 (19.2%) | |
| Other | 24 (7.8%) | 15 (7.4%) | 9 (8.7%) | |
| Missing | 9 (2.9%) | 9 (4.4%) | 0 (0.0%) | |
| Scleroderma subtype, n (%) | 0.71 | |||
| Limited | 183 (59.6%) | 117 (57.6%) | 66 (63.5%) | |
| Diffuse | 98 (31.9%) | 67 (33.0%) | 31 (29.8%) | |
| Unclassified | 18 (5.9%) | 11 (5.4%) | 7 (6.7%) | |
| Missing | 8 (2.6%) | 8 (3.9%) | 0 (0%) | |
| Antibody, n (%) | 0.64 | |||
| Anti-centromere | 81 (26.4%) | 50 (24.6%) | 31 (29.8%) | |
| Anti-topoisomerase (Scl-70) | 40 (13.0%) | 31 (15.3%) | 9 (8.7%) | |
| RNA polymerase III | 10 (3.3%) | 8 (3.9%) | 2 (1.9%) | |
| Isolated nucleolar ANA | 61 (19.9%) | 39 (19.2%) | 22 (21.2%) | |
| Anti-U1RNP | 14 (4.6%) | 9 (4.4%) | 5 (4.8%) | |
| Mixed or Other | 62 (20.2%) | 39 (19.2%) | 23 (22.1%) | |
| Negative | 20 (6.5%) | 14 (6.9%) | 6 (5.8%) | |
| Missing | 19 (6.2%) | 13 (6.4%) | 6 (5.8%) | |
| Digital ulcers, present, n (%)a | 10 (3.3%) | 7 (3.4%) | 3 (2.9%) | 1.00 |
| Modified Rodnan Skin Score, median (IQR) | 5 (3, 11) | 5 (3, 11.5) | 5 (3, 10) | 0.73 |
| Home oxygen use, n (%)a | 79 (25.7%) | 60 (29.6%) | 19 (18.3%) | 0.04 |
| Follow-up time, median (IQR) | ||||
| Index hospitalization to last PHAROS visit, median (IQR), years | 2.6 (1.3, 4.4) | 2.4 (1.0, 4.1) | 2.8 (1.7, 4.9) | 0.027 |
| Index hospitalization to next PHAROS visit, median (IQR), days | 64 (28, 123) | 81 (42, 136) | 37 (3, 80) | <0.001 |
| Index Hospitalization Reason, n (%) | ||||
| Related to PH | 113 (36.8%) | 59 (29.1%) | 54 (51.9%) | <0.001 |
| Related to SSc | 98 (31.9%) | 66 (32.5%) | 32 (30.8%) | |
| Not related to PH or SSc | 92 (30.0%) | 74 (36.5%) | 18 (17.3%) | |
| Missing | 4 (1.3%) | 4 (2.0%) | 0 (0.0%) | |
| PH type, n (%) | 0.02 | |||
| At riskb | 51 (16.6%) | 32 (15.8%) | 19 (18.3%) | |
| WHO group 1 | 124 (40.4%) | 79 (38.9%) | 45 (43.3%) | |
| WHO group 2 | 29 (9.4%) | 13 (6.4%) | 16 (15.4%) | |
| WHO group 3 | 26 (8.5%) | 22 (10.8%) | 4 (3.8%) | |
| Missing | 77 (25.1%) | 57 (28.1%) | 20 (19.2%) | |
| WHO functional class, n (%) | 0.25 | |||
| I/II | 191 (62.2%) | 121 (59.6%) | 70 (67.3%) | |
| III/IV | 107 (34.9%) | 75 (36.9%) | 32 (30.8%) | |
| Missing | 9 (2.9%) | 7 (3.4%) | 2 (1.9%) | |
| Pulmonary function tests, median (IQR) | ||||
| FVC, %-predicted | 76 (62.9, 89.9) | 78 (63.8, 91.8) | 73 (62.5, 86.3) | 0.22 |
| DLCO, %-predicted | 40 (30.4, 51.8) | 40 (30.2, 51.5) | 40 (31.0, 52.1) | 0.89 |
| TLC, %-predicted | 80 (63.6, 92.6) | 82 (62.4, 95.4) | 78 (66.0, 89.7) | 0.36 |
Gender missing for five (all without readmission); digital ulcers missing for four (three without and one with readmission); home oxygen missing for two (one with and one without readmission). bAt risk = high-risk for SSc-associated pulmonary hypertension with (i) systolic pulmonary arterial pressure >40 mmHg on echocardiography, (ii) %-predicted DLCO <55, or (iii) %-predicted FVC to DLCO ratio >1.6.
DLCO: carbon monoxide diffusing capacity; FVC: forced vital capacity; IQR: interquartile range; TLC: total lung capacity; WHO: World Health Organization.
Individuals who used (vs did not use) home oxygen were more commonly male (32% vs 15%, P =0.001), had higher prevalence of WHO group 3 PH (18% vs 5%, P =0.02), and had lower %-predicted FVC (72 vs 79, P <0.001), DLCO (32 vs 43, P <0.001) and TLC (72 vs 82, P =0.01) (Supplementary Table S2, available at Rheumatology online). According to home oxygen use, there was no significant difference in age, patient-reported race, antibody profile or scleroderma disease subtype.
Predictors of hospital readmission
The unadjusted logistic regressions evaluating predictors of hospital readmission are shown in Fig. 2. Reasons for readmissions included: PH-related (n = 54), SSc-related (n = 32) or neither PH- nor SSc-related (n = 18). Compared with index admissions related to neither PH nor SSc, individuals with index admissions related to SSc had 2-fold higher odds of readmission (OR 1.99; 95% CI 1.02, 3.87), and individuals with index admissions related to PH had approximately 4-fold higher odds of readmission (OR 3.76, 95% CI 2.00, 7.09) (Fig. 2).
Fig. 2 .
Risk factors for 12-month readmission among individuals with SSc-associated pulmonary hypertension, univariable analysis
Unadjusted odds ratios presented from univariable analysis. ‘Other’ as reason for index admission indicates those admission reasons related to neither PH nor SSc. SSc disease duration defined as time since first non-Raynaud syndrome. WHO: World Health Organization.
The multivariable logistic regression model evaluating predictors of hospital readmission is shown in Fig. 3. The odds of 12-month readmission after an index hospitalization for PH was further increased in models adjusting for WHO PH Group, age, sex, patient-reported race and home oxygen use (aOR 6.6; 95% CI 3.2, 13.6). Readmission was less likely among individuals who required vs did not require home supplemental oxygen, when adjusting for WHO PH Group, index hospitalization reason, age, sex and patient-reported race (aOR 0.44; 95% CI 0.22, 0.89) (Fig. 3). WHO group 3 was independently associated with decreased odds of readmission in the adjusted model (aOR 0.21; 95% CI 0.05, 0.80). In both unadjusted and adjusted models, patient-reported race, age and sex were not significantly associated with likelihood of readmission.
Fig. 3 .
Risk factors for 12-month readmission among individuals with SSc-associated pulmonary hypertension, multivariable analysis
Adjusted odds ratio presented in multivariable models adjusting for variables shown. ‘Other’ as reason for index admission indicates those admission reasons related to neither pulmonary hypertension (PH) nor SSc. SSc disease duration defined as time since first non-Raynaud syndrome. WHO: World Health Organization.
Sensitivity analysis
Overall, 31 of 307 ever-hospitalized individuals (10%) had a readmission within 90 days of an index hospitalization discharge, and 17 individuals (6%) had a readmission within 30 days of an index hospitalization discharge. Sensitivity analysis was performed for characteristics associated with readmission at 30 days and 90 days. Individuals with vs without 90-day readmission were less commonly classified has having Functional Class I (16% vs 29%; P =0.044). No patient-level factors remained statistically significant for 30-day readmissions.
Discussion
We sought to identify individual-level risk factors for 12-month hospital readmissions among individuals with SSc-PH. In this observational SSc-PH registry, 34% of individuals hospitalized during the study period experienced a readmission within 12 months. Readmission was more likely following hospital discharges for PH or SSc (compared with non-PH or SSc reasons) and less likely among individuals who use home oxygen or have WHO group 3 PH.
The observed 12-month readmission rate (34%) is lower than a previous study where 42% of ever-hospitalized individuals with PAH (of any cause) experienced readmission within 12 months [6]. The observed 30-day readmission rate (6%) is also lower compared with a previous study using the United States National Readmission Database where 22% of hospitalized individuals with SSc-PH experienced readmission within 30 days [5]. One possible reason for the higher readmission rates in previous studies is that those analysed larger, claims-based [6] or nationally representative [5] cohorts that incorporate individuals with potentially less access to specialized outpatient care. This finding underscores the importance of referral to specialized SSc and/or PH centres. Additionally, patient goals of care are not collected in these studies. These may differ between study cohorts and impact hospitalization rates.
The use of home oxygen was associated with a lower likelihood of readmission within 12 months in models that adjusted for PH type. These observational results cannot definitively characterize the relationship between home oxygen use and lower readmission rates. Potential reasons for this unexpected finding could be related to increased outpatient visits related to home oxygen use, increased use of therapies unmeasured in this study such as pulmonary rehabilitation, frequent assessments that allow for early home oxygen initiation within specialized centres, or a primary beneficial effect of oxygen on pulmonary disease and dyspnea. These findings add to results from the Registry to Evaluate Early and Long-term PAH Disease Management (REVEAL), a 5-year observational PAH study that includes individuals with SSc, where home oxygen use was associated with lower mortality among those with severely reduced %-predicted DLCO [19].
In adjusted models, the strongest factor associated with readmission was the index admission primary reason. If related to PH or SSc (vs unrelated to PH or SSc), odds of 12-month readmission increased significantly (aOR 6.6 and 2.2, respectively). This finding complements prior work demonstrating that hospitalizations related to PAH portended poorer survival at three years compared with those hospitalizations un-related to PAH [7]. A recent study using ICD-9 codes to characterize primary admission reason found lower risk for 30-day readmission among those whose primary index admission discharge diagnosis was pulmonary hypertension. This difference may relate to potential misclassification of documenting primary discharge using ICD-9 diagnosis codes. In contrast, the present study has highly characterized hospitalization reasons recorded by PHAROS investigators.
Among 29 individuals with WHO group 2 PH (related to systolic or diastolic heart failure), the majority [16 of 29 (55%)] experienced readmission within 12 months. This finding adds to prior work by Chung et al. that found CHF to be the third most common principal hospital diagnosis among individuals with SSc hospitalized between 2002 and 2003, using the National Inpatient Sample database [20]. However, we observed that when adjusting for age, race, sex, home oxygen use and index hospitalization reason, WHO group 2 was not independently associated with an increased odds of 12-month readmission. This finding underscores the morbidity related to PH that is distinct from that related to heart failure.
We observed no difference in readmission by scleroderma subtype. These data are in agreement with results of a 2013 meta-analysis where investigators found no significant survival difference between individuals with SSc-PH with limited vs diffuse disease (pooled hazard ratio 0.7; 95% CI 0.47, 1.03) [21]. We also did not observe differences in readmission by race. This was unexpected, because prior work has shown that African-Americans compared with Americans of European descent had more severe PAH [higher frequency of Functional Class III-IV (80% vs 53%; P =0.02), more right ventricular dysfunction, and higher brain natriuretic peptide (BNP) (5729 vs 1892 pg/ml; P =0.02)] [22]. Chung et al. also described that non-white race was associated with longer length of stay in individuals with SSc hospitalized for any reason [20]. Our study was not able to assess length of stay that could impact readmission risk. Additionally, our small sample size of predominantly white individuals may have limited our ability to assess readmission risk differences according to race.
We acknowledge our study limitations. Data were previously collected; thus, it was not possible to analyse some variables of interest such as hospital length of stay and discharge disposition. Additionally, missing data was present, particularly related to medication use. This limited our ability to evaluate the association between cyclophosphamide use and hospital readmission for study subgroups (such as those with and without interstitial lung disease). Due to the observational nature of the data, we are unable to draw causal inferences, and results may be influenced by unmeasured confounders (such as access to primary care). Further, analysed data were collected at encounter closest to hospitalization and may not accurately reflect the clinical condition at time of hospital discharge. This would most significantly impact the analysis of variables that can fluctuate significantly, such as patient-reported outcomes (vs age, disease subtype, antibody profile). Because the individuals included in this analysis were cared for at tertiary scleroderma centres, these results may not be generalizable to those with varying access to outpatient care for whom readmission risk may be higher. Also, the small sample size limits the assessment of patient-factors associated with 30- and 90-day hospital readmissions.
Study strengths include the use of a well-characterized cohort with prospectively collected data pertaining to PH and SSc disease severity. The reason for each hospital admission is also well characterized by investigators highly familiar with both SSc and PH. Additionally, PHAROS is a multi-site study, and the analytic cohort described herein includes individuals from 25 US sites. This enhances the generalizability of these results to other tertiary care centres.
In conclusion, among individuals with SSc-PH (or with SSc at high-risk for PH) cared for at tertiary centres, readmission within 12 months of a hospital discharge is common. Hospital discharges related to PH were associated with the highest odds of readmission within 12 months. This finding emphasizes the importance of outpatient clinical coordination following hospitalizations related to PH. We also found that home oxygen use was associated with lower odds of 12-month hospital readmission. Need for home oxygen can be routinely assessed at the time of hospital discharge in individuals with SSc-PH. Future prospective studies should determine whether universal testing for the need for home supplemental oxygen at the time of discharge reduces readmissions among those with SSc-PH.
Supplementary Material
Acknowledgements
We would like to acknowledge the PHAROS investigators: Alkassab F, Bolster M, Chung L, Csuka ME, Derk C, Domsic R, Fischer A, Frech T, Furst D, Goldberg A, Gomberg-Maitland M, Gordon J, Hant F, Hinchcliff M, Horn E, Hummers L, Hsu V, Kafaja S, Khanna D, Mayes MD, Medsger T, Molitor J, Preston I, Saketkoo L, Schiopu E, Shapiro L, Silver R, Simms R, Steen V, Varga J.
Funding: This study was an investigator-initiated study that has been supported by the Scleroderma Foundation, the Mackley Foundation, as well as by Actelion Pharmaceuticals and Gilead Sciences. This work was also supported by Weill Cornell Medicine Clinical and Translational Science Center (UL1 TR002384, K.S.).
Disclosure statement: K.S. reports no disclosures; L.C.P. reports no disclosures; D.J.-K. reports no disclosures; I.S. reports no disclosures; J.S. reports no disclosures; J.F. reports no disclosures; V.D.S. reports receiving funds for the following activities: Research Support: Reata Pharmaceuticals; Consulting: Johnson & Johnson, outside the submitted work; J.K.G. reports receiving funds for the following activities: Research Support: Corbus Pharmaceuticals, Cumberland Pharmaceuticals, and Eicos Sciences, outside the submitted work.
Data availability statement
Data are available upon reasonable request by any qualified researchers who engage in rigorous, independent scientific research, and will be provided following review and approval of a research proposal and Statistical Analysis Plan (SAP) and execution of a Data Sharing Agreement (DSA). All data relevant to the study are included in the article.
Supplementary data
Supplementary data are available at Rheumatology online.
Contributor Information
Kimberly Showalter, Department of Medicine, Division of Rheumatology, Hospital for Special Surgery.
Laura C Pinheiro, Department of Medicine, Division of General Internal Medicine, Weill Cornell Medicine.
Deanna Jannat-Khah, Department of Medicine, Division of Rheumatology, Hospital for Special Surgery.
Irina Sobol, Department of Medicine, Division of Cardiology, Weill Cornell Medicine.
Jackie Szymonifka, Department of Medicine, Division of Rheumatology, Hospital for Special Surgery.
Jackie Finik, Department of Medicine, Division of Rheumatology, Hospital for Special Surgery.
Virginia D Steen, Department of Medicine, Division of Rheumatology, MedStar Georgetown University Hospital, Washington DC, USA.
Jessica K Gordon, Department of Medicine, Division of Rheumatology, Hospital for Special Surgery.
References
- 1. Hao Y, Hudson M, Baron M et al. Early mortality in a multinational systemic sclerosis inception cohort. Arthritis Rheumatol 2017;69:1067–77. [DOI] [PubMed] [Google Scholar]
- 2. Steen VD, Medsger TA. Changes in causes of death in systemic sclerosis, 1972-2002. Ann Rheum Dis 2007;66:940–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Simonneau G, Galie N, Rubin LJ et al. Clinical classification of pulmonary hypertension. J Am Coll Cardiol 2004;43:5S–12S. [DOI] [PubMed] [Google Scholar]
- 4. Hinchcliff M, Fischer A, Schiopu E, Steen VD, Investigators P. Pulmonary Hypertension Assessment and Recognition of Outcomes in Scleroderma (PHAROS): baseline characteristics and description of study population. J Rheumatol 2011;38:2172–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Showalter K, Ma X, Pinheiro L et al. Thirty-day hospital readmission in systemic sclerosis associated pulmonary hypertension: a nationwide study. Semin Arthritis Rheum 2021;51:324–30. [DOI] [PubMed] [Google Scholar]
- 6. Burke JP, Hunsche E, Regulier E et al. Characterizing pulmonary hypertension-related hospitalization costs among Medicare Advantage or commercially insured patients with pulmonary arterial hypertension: a retrospective database study. Am J Manag Care 2015;21:s47–58. [PubMed] [Google Scholar]
- 7. Burger CD, Long PK, Shah MR et al. Characterization of first-time hospitalizations in patients with newly diagnosed pulmonary arterial hypertension in the REVEAL registry. Chest 2014;146:1263–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Coffey CM, Sandhu AS, Crowson CS et al. Hospitalization rates are highest in the first 5 years of systemic sclerosis: results from a population-based cohort (1980-2016). J Rheumatol 2021;48:877–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Piga M, Casula L, Sanna S et al. Population-based analysis of hospitalizations for patients with systemic sclerosis in a West-European region over the period 2001-2012. Rheumatol Int 2016;36:73–81. [DOI] [PubMed] [Google Scholar]
- 10.Preliminary criteria for the classification of systemic sclerosis (scleroderma). Subcommittee for scleroderma criteria of the American Rheumatism Association Diagnostic and Therapeutic Criteria Committee. Arthritis Rheum 1980;23:581–90. [DOI] [PubMed] [Google Scholar]
- 11. LeRoy EC, Black C, Fleischmajer R et al. Scleroderma (systemic sclerosis): classification, subsets and pathogenesis. J Rheumatol 1988;15:202–5. [PubMed] [Google Scholar]
- 12. Simonneau G, Montani D, Celermajer DS et al. Haemodynamic definitions and updated clinical classification of pulmonary hypertension. Eur Respir J 2019;53:1801913. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Hsu VM, Chung L, Hummers LK et al. Development of pulmonary hypertension in a high-risk population with systemic sclerosis in the Pulmonary Hypertension Assessment and Recognition of Outcomes in Scleroderma (PHAROS) cohort study. Semin Arthritis Rheum 2014;44:55–62. [DOI] [PubMed] [Google Scholar]
- 14. Ware JE Jr., Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care 1992;30:473–83. [PubMed] [Google Scholar]
- 15. Eakin EG, Resnikoff PM, Prewitt LM, Ries AL, Kaplan RM. Validation of a new dyspnea measure: the UCSD Shortness of Breath Questionnaire. University of California, San Diego. Chest 1998;113:619–24. [DOI] [PubMed] [Google Scholar]
- 16. Steen VD, Medsger TA Jr. The value of the Health Assessment Questionnaire and special patient-generated scales to demonstrate change in systemic sclerosis patients over time. Arthritis Rheum 1997;40:1984–91. [DOI] [PubMed] [Google Scholar]
- 17. White IR, Royston P, Wood AM. Multiple imputation using chained equations: issues and guidance for practice. Stat Med 2011;30:377–99. [DOI] [PubMed] [Google Scholar]
- 18. White IR, Carlin JB. Bias and efficiency of multiple imputation compared with complete-case analysis for missing covariate values. Stat Med 2010;29:2920–31. [DOI] [PubMed] [Google Scholar]
- 19. Farber HW, Badesch DB, Benza RL et al. Use of supplemental oxygen in patients with pulmonary arterial hypertension in REVEAL. J Heart Lung Transplant 2018;37:948–55. [DOI] [PubMed] [Google Scholar]
- 20. Chung L, Krishnan E, Chakravarty EF. Hospitalizations and mortality in systemic sclerosis: results from the Nationwide Inpatient Sample. Rheumatology 2007;46:1808–13. [DOI] [PubMed] [Google Scholar]
- 21. Lefevre G, Dauchet L, Hachulla E et al. Survival and prognostic factors in systemic sclerosis-associated pulmonary hypertension: a systematic review and meta-analysis. Arthritis Rheum 2013;65:2412–23. [DOI] [PubMed] [Google Scholar]
- 22. Blanco I, Mathai S, Shafiq M et al. Severity of systemic sclerosis-associated pulmonary arterial hypertension in African Americans. Medicine 2014;93:177–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data are available upon reasonable request by any qualified researchers who engage in rigorous, independent scientific research, and will be provided following review and approval of a research proposal and Statistical Analysis Plan (SAP) and execution of a Data Sharing Agreement (DSA). All data relevant to the study are included in the article.