The impact of pathways: a significant decrease in mortality

Abstract

This study was undertaken to determine how care pathways (CPs) in the hospital treatment of heart failure (HF) affected in-hospital mortality, and outcomes at discharge. A two-arm, cluster randomized trial was conducted. Fourteen community hospitals were randomized either to arm 1 (CPs) or to arm 2 (no intervention, usual care). A sample size of 424 patients (212 in each group) was used in order to have 80% of power at the 5% significance level (two-sided). The primary outcome measure was in-hospital mortality. Secondary outcomes were also evaluated. In-hospital mortality was 5.6% in the experimental arm and 15.4% in the controls (P = 0.001). In CP and usual care groups, the mean rates of unscheduled readmissions were 7.9% and 13.9%, respectively. Adjusting for age, smoking, New York Heart Association (NYHA) score, hypertension and source of referral, patients in the CP group, as compared with controls, had a significantly lower risk of in-hospital death (odds ratio [OR] = 0.18; 95% confidence interval [CI]: 0.07–0.46) and unscheduled readmissions (OR = 0.42; CI = 0.20–0.87). No differences were found between CP and control with respect to the appropriateness of the stay, costs and patient's satisfaction. This paper examines the evaluation of a complex intervention and adds evidence to previous knowledge, indicating that CP should be used to improve the quality of hospital treatment of HF.

Introduction

In Europe approximately 5% of all acute medical admissions relate to heart failure (HF), and in the United States HF is responsible for almost one million hospitalizations annually. Almost three-quarters of these admissions are unplanned and worsening HF is responsible for half of these admissions.^1,2

The Acute Decompensated Heart Failure National Registry (ADHERE) showed that the hospital treatment of HF frequently does not follow published guidelines nor does it conform to the Joint Commission on Accreditation of Healthcare Organizations (JCAHO) core performance measures, potentially contributing to the high morbidity, mortality and economic cost of this disorder.³ ADHERE findings also suggested that the wide variations in conformity may reflect differences in training, guideline familiarity, and implementation of tools and systems to ensure that recommended care is provided and documented. Consequently, the development of educational and quality improvement programmes has the potential to considerably reduce current variability in care, enhance guideline adherence and improve outcomes for patients.

Care pathways (CPs) have become a popular tool to achieve such goals.⁴ Despite enthusiasm and diffusion, the widespread acceptance of CPs remains questionable because very little prospective controlled data demonstrated their effectiveness.^5–8

The study was designed in order to conduct a rigorous evaluation of CPs in the hospital treatment of HF. The primary objective of the trial was to evaluate the effectiveness of the implementation of CPs for the hospital treatment of HF among a sample of Italian hospitals. Our hypothesis was that the CPs should be more effective than usual care in treating patients admitted to hospital for HF, and that the CPs they should reduce patients' mortality during the stay and they should improve patients' outcomes at discharge. Secondary objectives were to gather information about the effectiveness of cluster randomized-controlled trial (cRCT) study design when evaluating CPs.

Methods/design

Study design

A cluster randomized-controlled clinical trial to evaluate the effect of applying CPs to patients with HF was performed. The results obtained on treating patients with CPs and the results obtained on treating patients with the usual care were compared. As a CP is not a single intervention to be compared with a placebo but its eventual benefits come from a mix of complex actions that are implemented at the institutional level (appropriate use of practice guidelines and supplies of drugs and ancillary services, new organization and procedures, patient education, etc.), hospitals, rather than individual patients, were randomly assigned to either introduce the pathway or continue usual care.^9,10

Forty hospitals were invited to participate in the study. Eighteen hospitals showed interest in the implementation of the CP, and 14 community hospitals were selected and randomized. The selection was based on comparability of their location, patient population, facilities and teaching status.

Study evaluations

The primary outcome measure was in-hospital mortality. The secondary outcomes were length and appropriateness of stay, rate of unscheduled readmissions, customers' satisfaction and costs of treating the patients. The quality of care was assessed by monitoring the use of diagnostic and therapeutic procedures during hospital stay and by using key quality indicators at discharge, as reported in previous studies.^11–14

Study sample

The sample included all patients treated by the hospitals during the experimental period with a principal diagnosis of HF (all International Classification Diseases 9th revision Clinical Modification [ICD-9CM] codes included in 428.xx code). The sample size was defined in order to detect a statistical difference in the mortality rate. Based on this goal, a sample size of 424 patients (212 in each group) was required for the study to have 80% power at the 5% significance level (two-sided). The sample size calculation was performed according to standard criteria for cluster randomized trials. The sample size was adjusted using an inflation factor of 1.51 to account for the cluster randomization: seven clusters per trial arm, a cluster size of 30 patients and an intracluster correlation (ICC) of 0.018.^15–17 At baseline the comparability of the two groups at admission was verified by measuring patients' age, sex, co-morbidities, risk factors and symptom severity. Patients with a current acute myocardial infarction or unstable angina were excluded from the study.

Intervention

Each hospital was assigned one methodological leader who assisted local multidisciplinary teams in the development of pathways and in project implementation. The teams were formed on a voluntary basis, received a three-day training in the development of CPs and constructed the CPs over a six-month period. The teams analysed their care processes, conducted research for the best evidence and defined appropriate goals to satisfy the multidimensional needs of the patients. These results were detailed into protocols and documentation, including the sequence of events and expected progress of the patients over time.⁴ After validation of the pathways, each team educated its hospital staff on the use of the CP and also monitored the use of the pathway.

Data analysis

Data were prospectively collected by local staff both in interventions and in control groups (physician and nurses who were trained in two prestudy educational events). The analysis was performed by the research team using the Fisher's exact test and the Kruskal–Wallis test for categorical and continuous variables, respectively. Statistical significance will be defined as a two-sided P value <0.05. All analyses were intention-to-treat and were carried out using STATA statistical software, version 8.2 (Stata Corporation, College Station, TX, 2003).

Ethics

The project (ClinicalTrials.gov ID [NCT00519038]) was exempt from ethical clearance according to the Italian Ministry of Health law number (ex art. 12bis D.lgs 229/99).

Results

The final sample consisted of 429 patients (214 receiving CP and 215 usual care). An average of 407 (min. 194, max. 633) HF patients were treated annually in CP group hospitals and an average of 396 in control group hospitals (min. 147, max. 640). At admission, there were no significant differences between the groups (the following results are presented as CPs versus usual care, P value), according to gender (male 47.7% versus 51.2%, P = 0.5) and selected clinical conditions (hypertension 72.0% versus 74.9%, P=0.6; chronic obstructive pulmonary disease 24.3% versus 27.0%, P = 0.6; diabetes 19.2% versus 17.7%, P = 0.7; smoking 15.9% versus 14.4%, P=0.7), while the mean age was slightly higher in the experimental group (81.7 versus 79.7 years, P = 0.011). Although only patients with New York Heart Association (NYHA) III–IV scores or higher had to be admitted, a minority of patients with NYHA II scores were also admitted because of low socioeconomic status (7%).

The discharge status of all patients, according to the type of treatment received, and several outcomes of care are reported in Table 1. In-hospital mortality was 5.6% in the group receiving CP and 15.4% in the control group (P=0.001). At bivariate analysis, most of the secondary outcomes showed better results in the experimental groups.

Table 1

Outcome indicators during hospital stay in care pathway and usual care groups

	Care pathway (n = 214)	Usual care (n = 215)
Indicator	% (95% CI)	% (95% CI)	P
Discharge status – severity
Death (in-hospital mortality)	5.6 (2.5–8.7)	15.4 (10.5–20.2)	0.001
NYHA I	38.8 (32.3–45.3)	25.6 (19.7–31.4)	0.003
NYHA II	52.3 (45.6–49.0)	49.3 (42.6–56.0)	0.5
NYHA III	2.8 (0.6–5.0)	8.4 (4.7–12.1)	0.012
Transferred to another hospital	0.5 (0.0–1.4)	1.5 (0.0–3.0)	0.3
Appropriateness of the stay	76.2 (70.5–81.9)	72.1 (66.1–78.1)	0.3
Rate of unscheduled readmissions	7.9 (4.3–11.6)	13.9 (9.3–18.6)	0.053
	Mean (95% CI)	Mean (95% CI)
Length of stay in days	10.4 (9.6–11.0)	11.4 (10.5–12.3)	0.028
Cost of the admission, €	2125 (2054–2197)	2211 (2134–2289)	0.11
Patients' satisfaction score	8.50 (7.9–9.1)	8.14 (7.5–8.8)	0.5

NYHA = New York Heart Association

The results of the multivariable random-effect logistic models predicting in-hospital mortality and unscheduled readmissions are shown in Table 2. Adjusting for age, smoking, NYHA score, hypertension and source of referral (general practitioner or other), patients who were managed and cared for according to CPs, as compared with subjects in usual therapy, had a significantly lower risk of in-hospital death (odds ratio [OR] = 0.18; 95% confidence interval [CI]: 0.07–0.46) and unscheduled readmissions (OR = 0.42; CI: 0.20–0.87).

Table 2

Results of the random-effects logistic models predicting in-hospital mortality and unscheduled re-admission

	In-hospital mortality		Unscheduled readmissions*
Variables	OR (95% CI)	P	OR (95% CI)	P
Age (five-year increment)	1.08 (0.86–1.35)	0.5	0.83 (0.68–1.01)	0.063
Care pathway versus usual care	0.18 (0.07–0.46)	<0.001	0.42 (0.20–0.87)	0.020
Smoking	10.4 (3.72–29.4)	<0.001	6.26 (2.86–13.70)	<0.001
NYHA score IV versus other NYHA scores	56.3 (15.7–202)	<0.001	0.13 (0.04–0.39)	<0.001
Admission from general practitioner	0.36 (0.17–0.77)	0.009	–	–
Hypertension	–	–	7.83 (2.24–27.3)	0.001

*Excluding those who died during their first admission

With regard to the use of diagnostic procedures, echocardiography, oximetry and diuresis monitoring were much more frequently utilized for patients treated according to the CP as compared with controls (Table 3). Also all medications were administered more frequently in the experimental group, with the exception of diuretics and antiplatelet agents (Table 3). Similarly, at discharge compared with controls, the proportion of patients receiving left ventricular function assessment, advice/counselling on smoking cessation and written discharge instructions was significantly higher in the experimental group, while angiotensin-converting enzyme inhibitors at discharge were more frequently requested in the CP group; however, this difference was not significant (Table 3).

Table 3

Diagnostic procedures and use of medications during hospital stay and at discharge in the care pathway and usual care groups

Event	Care pathway (%)	Usual care (%)	P
Diagnostic procedure during hospital stay
Echocardiography	77.1	12.1	<0.001
Trans-oesophageal echocardiography	19.2	1.4	<0.001
Electrocardiography	96.3	95.3	0.6
Chest X-rays	90.6	79.1	0.001
Oximetry	73.8	26.0	<0.001
Weight monitoring	47.7	36.7	0.022
Diuresis monitoring	67.8	49.8	<0.001
Medical treatment during hospital stay
Inotropes	66.4	48.8	<0.001
ACE inhibitors	57.9	40.0	<0.001
Beta-blockers	46.7	10.2	<0.001
Diuretics	95.3	95.8	0.8
Nitrates	33.2	24.2	0.039
Other vasodilators	39.2	3.3	<0.001
Heparin	50.0	19.1	<0.001
Oral anticoagulants	58.9	14.9	<0.001
Antiplatelet agents	31.8	37.7	0.2
Evidence-based key interventions at discharge
Left ventricular function assessment at discharge or planned for after discharge	98.13	12.09	<0.001
Advice/counselling for smoking cessation	58.82	12.90	0.004
Written instructions at discharge addressing activity level, diet, discharge medications, follow-up, weight monitoring and what to do if symptoms worsen	61.68	8.84	<0.001
ACE inhibitor at discharge without contraindications, with LVF < 40%	60.0	41.7	0.23

LVF = left ventricular function; ACE = angiotensin-converting enzyme

Discussion

The main finding of this study is that care delivered using CPs based on current guidelines, as compared with usual care, is more effective in reducing in-hospital mortality of HF patients.

The observed reduction in hospital mortality may have some potential explanations, which are not mutually exclusive, and include improved performance, earlier discharge of palliative patients, and finally a baseline difference between the two groups in terms of vital signs, laboratory data and/or co-morbidities, due to selection bias. Concerning the latter, after randomization the two groups were similar for co-morbidities and NYHA scores. In addition, CP use remained a strongly significant determinant of mortality after adjusting for several potential confounders.

The second possible explanation for the observed findings is a different attitude between the groups towards the care of patients close to death. Indeed, if the hospitals in the experimental groups discharged patients when close to death, this would have artificially reduced their in-hospital mortality rate. However, although we did not have reliable data on the events after discharge, the average length of stay in CPs was only one day lower than in usual care and, more importantly, the severity at discharge according to NYHA classes was lower in the experimental group.

A significant improvement of most outcomes of care, including the rate of unscheduled readmissions, length of stay, diagnostic procedures and medication use, and most JCAHO quality indicators was also observed. With regard to secondary outcomes, the observed results demonstrated how evidence-based care can be effectively implemented in real-world patients hospitalized with HF. Indeed, the four JCAHO performance measures were improved in the CP group, in which we also detected a significantly better diagnostic and therapeutic pattern. These findings were observed despite the complexity of an objective evaluation of the processes of care because of dependence on clinical practice, susceptibility to inadvertent bias and lack of validated measures.

Because of the impact of the results, some comments about the effectiveness of the study design are needed. A typically controlled trial design is not used in evaluating CPs because the context level adaptation, which is essential for pathways to work, is perceived as inappropriate in the trial design, likely because of the difficulty of keeping the intervention replicable and recognizable.¹⁸ To this purpose we defined as standard the steps in the change process or the key functions that the elements of the intervention were meant to improve according to each context. Also the definition of the quality care indicators helped. The indicators were driven by the theory and concerned the functions provided by the key elements of the intervention that were based on expected adherence to the same evidence. We think that this strategy based on combining local change standards with the use of shared evidence-based indicators maintained the integrity of the intervention in each site.

As reported in sample size literature in a cRCT, we had difficulties in defining the sample size for the dual nature of the trial that focused on both individuals and clusters. The sample size calculation was based on the number of individuals needed, while the randomization process was based on clusters. As each additional cluster represented a large proportionate increase in study size (and in associated costs), it was necessary to find the proper balance between the need to increase the number of clusters and organizational costs. According to cluster design, the sample size calculation needed to predict not only the expected effect size but also the anticipated cluster size and ICC, and this was difficult because of the lack of published data on CPs. However, we think that the number of clusters and of individuals included in our sample ensured the viability of the trial.^19,20 Given the type of intervention, a cluster randomized design was probably the most appropriate design to use.^10,17

In light of the above considerations, it is reasonable to believe that the observed reduction in hospital mortality should be attributed to an improvement in the quality of care determined by the use of CPs. In conclusion, the present study adds to the evidence, suggesting that the implementation of CPs for the hospital treatment of HF produces a relevant improvement in the quality of care and reduces in-hospital mortality. Further efforts should be devoted to enhance the diffusion of CPs based on reliable evidence and to define a better framework in which to evaluate CPs as a complex intervention.

References

Cowie

, Fox

, Wood

, Hospitalization of patients with heart failure. A population-based study. Eur Hearth J 2002;11:877–85

American Heart Association American Stroke association. Hearth Disease and Stroke Statistics. See http://www.americanheart.org/downloadable/heart/1105390918119HDSStats2005Update.pdf (last accessed 21 September 2009)

Fonarow

, Yancy

, Heywood

. ADHERE Scientific Advisory Committee, Study Group, and Investigators. Adherence to heart failure quality-of-care indicators in US hospitals: analysis of the ADHERE Registry. Arch Intern Med 2005;165:1469–77

European Pathways Association (EPA). Clinical/Care Pathways. See http://www.e-p-a.org/000000979b08f9803/index.html (last accessed 21 September 2009)

Panella

, Marchisio

, Di Stanislao

. Reducing clinical variations with clinical pathways: do pathways work? Int J Qual Health Care 2003;15:509–21

Rotter

, Kugler

, Koch

, A systematic review and meta-analysis of the effects of clinical pathways on length of stay, hospital costs and patient outcomes. BMC Health Serv Res 2008;8:265

Lemmens

, van Zelm

, Vanhaecht

, Kerkkamp

. Systematic review: indicators to evaluate effectiveness of clinical pathways for gastrointestinal surgery. J Eval Clin Pract 2008;14:880–7

Barbieri

, Vanhaecht

, Van Herck

, Effects of clinical pathways in the joint replacement: a meta-analysis. BMC Med 2009;7:32

Campbell

, Fitzpatrick

, Haines

, Framework for design and evaluation of complex interventions to improve health. Br Med J 2000;321:694–6

10.

Loeb

. Application of the development stages of a cluster randomized trial to a framework for valuating complex health interventions. BMC Health Serv Res 2002;2:13–21

11.

Bonow

, Bennett

, Casey

Jr , American College of Cardiology, American Heart Association Task Force on Performance Measures (Writing Committee to Develop Heart Failure Clinical Performance Measures), Heart Failure Society of America. ACC/AHA clinical performance measures for adults with chronic heart failure: a report of the American College of Cardiology/American Heart Association Task Force on Performance Measures (Writing Committee to Develop Heart Failure Clinical Performance Measures) endorsed by the Heart Failure Society of America. J Am Coll Cardiol 2005;46:1144–78

12.

Spertus

, Eagle

, Krumholz

, Mitchell

, Normand

. American College of Cardiology and American Heart Association methodology for the selection and creation of performance measures for quantifying the quality of cardiovascular care. J Am Coll Cardiol 2005;45:1147–56

13.

Joint Commission on Accreditation of Healthcare Organizations. Specifications Manual for National Hospital Quality Measures 2006. See http://www.jointcommission.org/PerformanceMeasurement/PerformanceMeasurement/Historical+NHQM+manuals.htm (last accessed 21 September 2009)

14.

Panella

, Marchisio

, Gardini

, Di Stanislao

. A cluster randomized controlled trial of a clinical pathway for hospital treatment of heart failure: study design and population. BMC Health Serv Res 2007;7:179

15.

Ukoumunne

, Gulliford

, Chinn

, Sterne

JAC

, Burney

PGJ

. Methods for evaluating area-wide and organization-based interventions in health and health care: a systematic review. Health Technol Assess 1999;3:iii–92

16.

Campbell

, Mollison

, Grimshaw

. Cluster trials in implementation research: estimation of intracluster correlation coefficients and sample size. Stat Med 2001;20:391–9

17.

Medical Research Council. Cluster Randomized Trials: Methodological and Ethical Considerations. London: Medical Research Council, 2002

18.

Hawe

, Shiell

, Riley

. Complex interventions: how ‘out of control’ can a randomised controlled trial be? Br Med J 2004;328:1561–3

19.

Campbell

, Donner

, Klar

. Developments in cluster randomized trials and Statistics in Medicine. Stat Med 2007;26:2–19

20.

Donner

, Klar

. Pitfalls of and controversies in cluster randomization trials. Am J Public Health 2004;94:416–22