Planned Re-admissions and Survival among Patients with Chronic Congestive Heart Failure - Essay Example

Critical Appraisal of the Article: Stewart, S., Marley, J.E., & Horowitz, J.D. (1999). Effects of a multidisciplinary, home-based intervention on planned readmissions and survival among patients with chronic congestive heart failure: a randomised controlled study. Lancet, 354, 1077-83. Introduction The purpose of this study by Stewart and colleagues (1999) was to measure the effects of a multidisciplinary home based intervention on planned readmissions and survival among patients with chronic congestive heart failure (CHF). The hypothesis (null) in this study was that there would be no difference in the frequency of unplanned readmission plus out-of hospital deaths, during a minimum of 6 months of follow-up, among patients with chronic CHF discharged home after acute admission between those exposed and not exposed to a CHF-specific, multidisciplinary, home based intervention in addition to usual care. The effects of home based intervention (HBI) in chronic illness are a much debated topic. The reason being that those who suffer from chronic illness like congestive heart failure, diabetes, chronic obstructive lung disease, etc., also contribute to a great deal of health care costs, mainly due to repeated admissions. Also, these patients are at a threat to sudden and emergency admissions, putting a risk on their lives. Many studies have been done to find out if home-based interventions by nursing personnel does help in bringing down hospital admissions and emergency admissions and improve patient out come. This research aims to study on this aspect in chronic heart failure patients. The Study This study included 200 patients from the inpatient department in a tertiary care hospital under the care of a cardiologist. The inclusion criteria were age- 55 years or more, about to be discharged home, congestive heart failure with left ventricular ejection fraction less than 55% and New York Heart Association Class 2, 3 or 4 and at least one previous admission for acute heart failure. The exclusion criteria were extensive, reversible ischaemia precipitating heart failure, valvular heart disease amenable to surgical correction, intended heart transplantation, presence of terminal disease, or home address outside the hospital’s catchment area.     To arrive at this study number, the researchers screened initially 4055 cardiology inpatients and narrowed down to 285 patients and finally to 200 patients after consent. The study was approved by ethics committee and consent of all participants also was taken.      The study was a randomised control study. Randomisation was initiated after measuring mental acuity, function status and extent of co-morbidity. 2 groups were formed. One group were subjected to usual care alone and the other to multidisciplinary home based intervention along with usual care. For both the groups the discharge planning was the same. This was adopted because; the authors postulated that home visits represent the most effective component of this type of intervention. Keeping in view that non-specialist management and inappropriate pharmacotherapy are associated with poorer health outcomes, all study patients were managed by a cardiologist and received treatment according to current guidelines.      The intervention used was Home Based Intervention (HBI) where in qualified cardiac nurse visited the patient at home after 7 to 14 days after discharge. The purpose of delaying the home visit was to allow potential problems to emerge which could be addressed. A specialist cardiac nurse was chosen for 2 reasons. Firstly, employing nurse is a much cost effective decision. Secondly, CHF patients need better assessment of their management of illness and a more accurate determination of their future needs which needs a lot of time and requires expertise in the field. During the visit, the nurse assessed the patients clinical progress since discharge. Physical examination, patient’s adherence to the prescribed treatment regimen, understanding of the disease process (including the ability recognise changes in symptoms indicative of worsening heart failure), fluid and sodium intake, current amount of physical activity, extent of psychosocial support, and use of available community-based resources, all were looked for. Based on these findings, patients and their families received appropriate instructions. A report of these findings and instructions was formed and sent to the patient and the cardiologist. These patients were contacted on phone at 3-6 months interval and home visits were repeated only if needed. The patient was encouraged to contact the cardiology nurse in case of any doubts. The outcome measures of this study were frequency of unplanned readmission and out of hospital deaths up to 6 months of follow up. Others outcomes which were looked for are event-free survival, cost of health care, longer term effects and health related quality of life within 6 months of index admission. Analysis Statistical analysis was done. SPSS for Windows (version 8.0) soft ware was used. For discrete variables, endpoint data were compared by means of x2 analysis (with calculation of odds ratio and 95% CI). For normally distributed continuous variables, end point data were compared with Student’s t test. For non-normally distributed continuous variables, Mann-Whitney test was used. For event -free survival and death, Kapalr -Meler survival curves were used. This was followed by the log-rank test for any difference between groups in the number and Breslow’s test to look for any difference between groups in the timing of events. To study the interaction between treatment mode and other potential correlates of the primary endpoint and overall mortality at 6 months after index admission, Cox proportional-hazards model was used. All analyses were done with an intention to treat. Results Though 200 patients were recruited, with 100 in each group, the intervention group finally had only 88 participants. This is because 2 died within 48 hours of discharge and 10 refused participation after consent. 40% of the intervention group had signs and symptoms of early clinical deterioration. There was no drug compliance in 25%. 90% were not aware of the importance of salt restriction. 48% were introduced to flexible diuretic regime based on daily weight checking. 38% needed physician review after recommendation by the specialist nurse. 22% needed the services of greater pharmacist contact. Home-support services were increased to 26%. 25% utilised the facility to contact the nurse in case of help. 4% of the routine phone calls made by the nurse turned in to immediate referral to physician. The main end-points measured were out of hospital deaths and readmissions during the 6 month follow up. There were 9 out-of hospital deaths in the intervention group and 11 in the usual-care group. There were 68 unplanned admissions in the intervention group and 118 in the usual care group (p= 0.031). Beyond 6 months, both the groups had similar admissions. There was 40% reduction in the number of primary events with the home-based intervention during the 6 month follow up (p= 0.02). The duration of event free survival persisted up to 9 months in the intervention group (p= 0.037). The intervention group had a lower number of unplanned days in hospital than the usual care group (p= 0.01). But they also had more elective days in hospital (p=0.13) mainly for surgical procedures. Regression analysis showed that the study was a border line independent predictor of survival at 6 months. As far as individual health care costs were considered, these were lower in the intervention group than the usual-care group (p= 0.16). This difference was largely due to lower recurrent unplanned admissions in the intervention group. The Quality of Life scores, in general were better in the intervention group. However, in those who survived the 6 months study period; these scores were similar to both the groups. The authors reported in their results that an inexpensive, non-pharmacological intervention improves the efficacy of pharmacotherapy in limiting readmission to hospital and death in a group of patients with severe CHF over a period of at least 6 months. They also opined that drugs alone do not have an impact on general poor health outcomes in CHF patients. They also require non-drug regimens. This is evident from the statistics. First of all, 25% of HBI patients were non-compliant to drugs. Also, 90% of patients were unaware of the importance of restricting salt intake. 48% patients needed to be changed to a flexible diuretic regime. HBI also is associated with better compliance and fewer medication-related admissions. The nurse was able to provide a better picture of clinical assessment because of longer duration of contact. Thus better care was possible. Also, in the process, the patients were more educated to manage themselves. The limitations of this study were: 1. Identification of exact mechanisms of effect of multifaceted interventions is difficult. 2. The delay in the first visit was indented for potential problems to arise so that they could be addresses. However 10% of the patients died during this phase. Hence delay in the first visit can be considered inappropriate. The timing of the home visit remains difficult. 3. The intensity of routine reinforcement required to sustain the efficacy of initially implemented strategies remains uncertain. 4. Intervention- group patients received more comprehensive medical care because treating physicians were aware of their status. 5. The findings of this study were conducted in a single health care system and hence may not be applicable to other health care systems. Critical review This is a good-size study. Actually the researchers’ previous study had pointed that 90 patients in each group would be required to detect a 20% difference in the end point (Stewart, 1998). This study included 100 in each group. In this study, the patients were randomised. Randomisation was concealed. The investigator was unaware of the demographic and clinical profile and was allocated the individual via phone by a separate unit. Equal size of groups and distribution of patients was done through a computer generated protocol. Patients were analysed in the groups to which they were randomised. The demography profile included sex, mean age, whether living alone, primary language and number of years of formal education ( Read More