Impact of Home Return Assistance Program (prado-ic) on the Re-hospitalisation Rate and Mortality for Hospitalized Heart Failure Patients

Citation:

Abstract

Keywords:

Heart Failure; Mortality; Congestive heart failure

Introduction

Congestive heart failure (CHF) is a common disease with poor outcome, and recurrent hospitalizations [1]. Therefore, patient journey is a key point: among several factors, transition of care must be improved [2,3].

The Caisse Nationale d'Assurance Maladie (CNAM, French National Social Security Service) has provided since 2013 a program named PRADO-IC (Programme d'Aide au Retour a Domicile-Insuffisance cardiaque, Returning Home Support Program- Heart Failure) to improve the transition of care. The care of the outpatient is based on his general practitioner and a private nurse [4]. Saint Joseph Hospital (SJH), is the first contributor of PRADO-IC program in the Ile-de-France Region [5].

The aim of this study is to compare the one-year mortality and the re-hospitalization rate in cohort of patients included in PRADO-IC versus a cohort not included (no PRADO-IC).

Patients and Methods

Population: From September 2016 to September 2018 all patients hospitalized for CHF in SJH were included. CHF was identified firstly by the appropriate WHO-ICD-10 (World Health Organization – International Classification of Diseases).

Two groups were compared according to the inclusion in PRADO-IC or not (Control group (C), the inclusion in PRADO group was at the physician’s discretion.

Data collection: Data were collected from the electronic health record.

Mortality and date of death if occurred were obtained using open-source national data base INSEE.

Primary endpoints: the rate of death, and of re-hospitalization for CHF at one year.

Statistical analysis: All quantitative variables were expressed as means ± standard deviations or median and interquartile range according to the data’s distribution; qualitative variables were expressed as number (percentages). Comparisons between quantitative or qualitative variables were made using appropriate tests.

Survival curves of patients and control were depicted by the Kaplan Meier method and compared using the Cox proportional hazard model.

All tests were two-sided, and a p value < 0.05 was considered as statistically significant.

Ethical considerations: This retrospective study was approved by the clinical ethic committee of the Groupe Hospitaller Paris Saint-Joseph (IRB 00012157) and was registered in ClinicalTrials.gov (NCT04613973).

Results

Selection of patients: According ICD-10 codes, 866 patients were selected. After application of ESC criteria [6], and exclusion for non-consent or discharge reasons, the study population were 633 patients, 262 in PRADO-IC group, and 371 in C group.

Table 1: Baseline characteristics of the study population are presented.

 

	Total (N=633)	Control (N=371)	PRADO (N=262)	p
Sex
Female	293 (46.3%)	158 (42.6%)	135 (51.5%)	0.0263
Male	340 (53.7%)	213 (57.4%)	127 (48.5%)
Age (years)
Mean (SD)	78.3 (11.1)	76.8 (11.6)	80.3 (10.1)	0.0001
LV EF (%)
Mean (SD)	45.2 (15.0)	45.3 (15.3)	45.0 (14.6)	0.86
LV EF
≤30%	153 (24.2%)	92 (24.8%)	61 (23.3%)	0.49
30%-50%	182 (28.8%)	100 (27.0%)	82 (31.3%)
≥50%	298 (47.1%)	179 (48.2%)	119 (45.4%)
Diabetes
No	427 (67.5%)	254 (68.5%)	173 (66.0%)	0.52
Yes	206 (32.5%)	117 (31.5%)	89 (34.0%)
Rhythm
Sinus rhythm	283 (44.7%)	180 (48.5%)	103 (39.3%)	0.0123
Atrial Fibrillation	263 (41.5%)	151 (40.7%)	112 (42.7%)
Pace maker	87 (13.7%)	40 (10.8%)	47 (17.9%)
Hypertension
No	165 (26.1%)	106 (28.6%)	59 (22.5%)	0.09
Yes	468 (73.9%)	265 (71.4%)	203 (77.5%)
BNP at discharge (ng/ml) (N=631)
Median [Q1;Q3]	352.0 [174.5;759.0]	316.0 [151.0;675.0]	396.0 [201.5;852.5]	0.0005
Plasmatic concentration of creatinine at discharge (µmol/l)
Mean (SD)	132.9 (62.5)	128.0 (63.1)	139.8 (61.1)	0.0196
Hemoglobin at discharge (g/dl)
Mean (SD)	12.4 (1.8)	12.7 (1.9)	12.0 (1.7)	<0.0001
Length of stay (days) (N=187)
Median [Q1;Q3]	6.0 [3.5;10.0]	6.0 [3.0;10.0]	7.0 [4.0;10.5]	0.29
Mortality rate at 1 year
No	522 (82.5%)	311 (83.8%)	211 (80.5%)	0.28
Yes	111 (17.5%)	60 (16.2%)	51 (19.5%)
Re-hospitalization rate for HF at 1 year
No	437 (69.0%)	267 (72.0%)	170 (64.9%)	0.06
Yes	196 (31.0%)	104 (28.0%)	92 (35.1%)
Time until readmission (for HF) (days) (N=196)
Median [Q1;Q3]	66.0 [23.0;190.8]	54.5 [21.8;201.5]	74.5 [24.0;181.5]	0.56

Comparison PRADO-IC group versus C group: PRADO-IC group is older and more often female, presented more often atrial fibrillation, high level of BNP or serum creatinine. The duration of in-hospital stay was not different.

The one-year mortality of the study population was 17.5% (n = 111), 19.5% (n = 51) in PRADO-IC group and 16.2% (n = 60) in C group, without significant difference (p = 0.28). Figure 1 shows the cumulative probability of survival in patients PRADO versus control.

Figure 1: Kaplan-Meier curves comparing all-cause mortality incidence.
Note: HR: Hazard Ratio [95%CI], P-value obtained by a Cox proportional hazard model.

The one-year re-hospitalization rate for CHF was 31.0% (n = 196) for the study population, 35.1% (n = 92) in PRADO-IC group, 28.0% (n = 104) in C group, without statistical difference (p = 0.006). Time to re admission was comparable between the two groups. Figure 2 shows the cumulative probability of re admission in the PRADO group and the Control group.

Figure 2: Kaplan-Meier curves comparing re-hospitalization incidence. Note: HR: Hazard Ratio [95%CI], P-value obtained by a Cox proportional hazard model.

Discussion

Clinical and biological characteristics of PRADO group are associated in the literature with a poor prognosis [7,8]. These differences showed an indication bias.

However, the lack of over-mortality in this high-risk group could be an indirect proof of the benefit of PRADO.

The lack of difference between the two groups concerning re-admission rate, which could be explained [9]. Only the high intensity programs (home visits with telephone follow up, clinical visits or both) reduced readmission risk. PRADO is a moderate intensity program.

Our investigation has few limitations. It is an observational, monocentric, real-world study, so its results may be influenced by potential information, selection and confusion biases, as previously described.

Our study is the largest cohort and longest duration of follow up that explored impact of PRADO-IC program. Moreover, the diagnosis of HF is specific, based on ESC criteria and not only ICD-10 codes. A further randomized prospective study is needed to assess the impact of PRADO-IC.

Conclusion

Our retrospective observational study confirms that the hospitalized patients for CHF are elderly patients with several comorbidities, high level of mortality and re admission rate at one year. Inclusion in PRADO-IC program concerned most severe patients. Despite this, PRADO-IC program provides a return-to-home ability and home care during 3 months after discharge, without a longer in-hospital stay, an over-mortality and re-admission occurrence while this population is exposed to these risks.

Disclosure of interest

The authors declare that they have no competing interest.