Abstract
OBJECTIVE To evaluate a new program, Integrating Physician Services in the Home (IPSITH), to integrate family practice and home care for acutely ill patients.
DESIGN Causal model, mixed-method, multi-measures design including comparison of IPSITH and non-IPSITH patients. Data were collected through chart reviews and through surveys of IPSITH and non-IPSITH patients, caregivers, family physicians, and community nurses.
SETTING London, Ont, and surrounding communities, where home care is coordinated through the Community Care Access Centre.
PARTICIPANTS A total of 82 patients receiving the new IPSITH program of care (including 29 family physicians and 1 nurse practitioner), 82 non-randomized matched patients receiving usual care (and their physicians), community nurses, and caregivers.
MAIN OUTCOME MEASURES Emergency department (ED) visits and satisfaction with care. Analysis included a process evaluation of the IPSITH program and an outcomes evaluation comparing IPSITH and non-IPSITH patients.
RESULTS Patients and family physicians were very satisfied with the addition of a nurse practitioner to the IPSITH team. Controlling for symptom severity, a significantly smaller proportion of IPSITH patients had ED visits (3.7% versus 20.7%; P = .002), and IPSITH patients and their caregivers, family physicians, and community nurses had significantly higher levels of satisfaction (P < .05). There was no difference in caregiver burden between groups.
CONCLUSION Family physicians can be integrated into acute home care when appropriately supported by a team including a nurse practitioner. This integrated team was associated with better patient and system outcomes. The gains for the health system are reduced strain on hospital EDs and more satisfied patients.
When patients are acutely ill, there are 6 main care alternatives available to them: 1) hospitalization; 2) emergency department (ED) use; 3) formal hospital-in-the-home programs; 4) office or home visits with family physicians; 5) community home care agencies; and 6) telehealth. Hospitalizations are infrequent and of short duration for acute illnesses for a variety of reasons. Hospital-in-the-home programs are not widely available and do not normally include family physicians, yet patients prefer them to hospitalization.1–3 The usual alternatives to hospitalization in Canada are office care by family physicians, ED use, telehealth, and community home care. Despite the fact that patients might be cared for by both family physicians and community care providers, these providers usually operate with no formal links between them.4,5
Formal hospital-in-the-home programs have been widely studied.1,6–15 These studies demonstrate equivalent outcomes to hospitalization, that hospital-in-the-home models are acceptable to patients, and mixed results in terms of cost-effectiveness and the acceptability of hospital-in-the-home programs to family caregivers. There is also work concerning case management in the community for long-term conditions.16–19 However, community-based care for acute illness that incorporates the patient’s family physician as part of a multi-disciplinary team is virtually unstudied. We do not know which community-based models of care for seriously ill adult patients work best for all participants: the patient, the family caregiver, the family physician, the nurses, and the system.
The evaluation reported here addressed this gap by quantitatively describing the implementation and process goals (such as participation, workload, and satisfaction) of a novel care program and by evaluating patient outcomes in a comparison study. Patients with acute or complex illness received enhanced home-based care from a multidisciplinary team including their own family physicians in a novel care program called Integrating Physician Services in the Home (IPSITH). The evaluation of this new IPSITH program using a causal model, mixed-method, multi-measures design20–26 was undertaken with an assessment of the perspective of all players. The qualitative component is reported elsewhere.27,28
METHODS
Setting
The setting was London, Ont, and its surrounding communities, where home care was coordinated through a regional provincially funded agency called Community Care Access Centres of London and Middlesex (CCAC). The program was conducted from 2000 until 2002.
Program and intervention
The IPSITH program was developed and implemented by a team of stakeholders including health care providers, researchers, and policy makers. A medical coordinator (family physician, 1 day per week) and a full-time nurse practitioner were hired. Family physicians were invited to participate in recruitment meetings, where the IPSITH program was described and patient eligibility was explained. A medical infrastructure was put in place within the existing local CCAC that sought to enhance the usual care provided.
Usual care included a case manager, who ordered nursing and allied health professional services as required. Patients receiving usual care could be admitted to home care by family physicians or home care professionals, or as early discharges from hospital.
The IPSITH program included the usual care providers with the addition of the patients’ family physicians, the IPSITH nurse practitioner, and in most cases, a family member or friend who acted as a caregiver in the home. There were 44 family physicians who attended the recruitment meetings and enrolled in the program, indicating a willingness to admit patients to IPSITH. Of these, 29 physicians enrolled patients into the program. Additionally, relationships and processes were established. A pharmacy, oxygen suppliers, and diagnostic services agreed to provide their services in patients’ homes, and 39 specialists agreed to provide urgent consultation for IPSITH patients upon request, with 18 agreeing to see patients in their homes if requested. Processes were established for the initial assessment, out-of-hours coverage, a record system, rapid response to crises, and fast-track admissions to hospital when needed. Most medical care services were coordinated through the nurse practitioner, who maintained close communication with patients’ family physicians. Family physicians made home visits as deemed necessary.
Patient participants
Eighty-two patients were enrolled consecutively in the IPSITH project. The IPSITH family physicians determined their patients’ eligibility for the program and admitted all patients deemed eligible with acute or complex conditions requiring care for an anticipated 5 to 10 days. Patients were asked to consent to the research component of the IPSITH project. For each IPSITH patient, a comparison (non-IPSITH) patient receiving usual acute care through CCAC was identified (matched on diagnosis and age within 10 years) by his or her CCAC case manager and asked to participate in the study. Both IPSITH and non-IPSITH physicians were asked to complete a Duke Severity of Illness Checklist29 for each patient to determine whether the IPSITH and non-IPSITH patients were adequately matched on disease symptom severity.
Data sources
The main sources of data were chart reviews, patient and caregiver interviews, and family physician and community nurse surveys. Survey response rates for the various participant groups were very high (75% to 100%), with the exception of non-IPSITH physicians (62.2%).
Chart reviews
Charts were reviewed to identify whether process goals were met, the number of physician home visits, and the number of telephone calls.
Patient and caregiver interviews
Interviews were conducted in the home within 2 weeks of discharge, including questions about patient and caregiver satisfaction (adapted from Richards et al30), patient and caregiver future preferences (one adapted from Richards et al30), patient ED use (adapted from Browne et al31,32), and caregiver burden.33
Family physician surveys
Surveys were mailed to participating physicians to collect information about patients’ disease symptom severity on a 5-point scale on the 4 dimensions of symptoms, complications, prognosis, and treatability (Duke Severity of Illness Checklist29), family physician satisfaction (1 item), barriers to providing home care (1 item), and attitudes about the relative quality of home and hospital care (1 item).
Community nurse surveys
Surveys were mailed to community nurses to collect information about nurse satisfaction (1 item) and attitudes about the relative quality of home and hospital care (1 item).
Analysis
The process evaluation describes the implementation of the IPSITH program through means and frequencies.
The non-randomized comparison of the IPSITH and non-IPSITH patient groups was conducted on the following outcomes: ED visits during the treatment period; ED visits during the 6 weeks following the treatment period; satisfaction (patient, caregiver, family physician, and nurse); caregiver burden; and patient and caregiver preferences for future care. Possible covariates (patient age, sex, diagnostic category, symptom severity, education, and income) were assessed for inclusion in the statistical models. Only symptom severity was found to vary between the IPSITH group and the non-IPSITH group, and thus was controlled for in all analyses. Analysis of variance was employed for continuous outcomes and logistic regression for dichotomous outcomes, controlling for the differences in symptom severity. Only 51 of 82 non-IPSITH family physicians returned their surveys including the Duke Severity of Illness Checklist. Of these, a further 15 indicated that they were unable to complete the Duke Severity of Illness Checklist because they did not have contact or involvement with their patients during the treatment episodes. This meant that for the non-IPSITH group, severity data were only available for 36 patients. The difference in severity between IPSITH and non-IPSITH patients for the remaining cases was imputed. The first step in imputation involved the observation that the difference in severity was related to diagnostic category. The second step involved assigning the median difference-in-severity scores from the actual data for each diagnostic category. In the third step, results were calculated with both the imputed symptom severity data (82 IPSITH and 82 non-IPSITH patients) and the actual severity data (36 IPSITH and 36 non-IPSITH patients).
Ethics approval, confidentiality, and security
The study received ethics approval from The University of Western Ontario Health Sciences Research Ethics Board. All data were kept strictly confidential.
RESULTS
Patient characteristics
Table 1 shows that there were no important demographic differences between the IPSITH patients and the non-IPSITH patients. Box 1 describes an example IPSITH case.
Patient characteristics: A) Age and length of treatment; B) Diagnostic category, symptom severity, education, and income.
Physician characteristics
The IPSITH physicians (n = 29) and non-IPSITH physicians (n = 69) did not vary by sex, year of graduation, hospital privileges, solo or group practice, or urban or rural practice. They did vary by membership in the College of Family Physicians of Canada, with more IPSITH physicians being members (85.7% vs 55.0%; P = .008). They also varied by teaching status, with more IPSITH physicians teaching (55.1% vs 5.9%; P < .001).
Process evaluation: IPSITH program
Key process goals for the development of the IPSITH program were met. One hundred percent of IPSITH patients’ charts indicated that there had been prompt initial medical assessment and treatment, a written discharge plan, and physician communication with community providers. Table 2 shows the number of family physicians by the number of cases admitted. Table 3 describes the workload for IPSITH family physicians.
Number of IPSITH family physicians by number of IPSITH cases admitted: The median (IQR) number of cases per family physician was 2 (1–3).
Workload for the 29 IPSITH family physicians: A) Mean (range) measures of workload elements per patient (n = 82); B) Time of day of housecalls.
Most patients and family physicians were satisfied with the addition of the nurse practitioner to the home care team (Table 4). The nurse practitioner made an average of 4.96 visits and a median of 4 visits per patient.
Patient and family physician satisfaction with the NP*
Figure 1 shows patients’ and caregivers’ preferences for location of care (home vs hospital); Figure 2 illustrates family physicians’ and community nurses’ assessment of quality of care (home vs hospital). Most patients, caregivers, and nurses reported that home care was strongly preferred to or much better than hospital care, compared with a minority of family physicians.
Patient and caregiver preferences for location of care (home vs hospital): Patients and caregivers were asked: “For the same condition in the future, where would you prefer treatment?”
Family physician and community nurse preference for location of care (home vs hospital): Family physicians and community nurses were asked: “Do you think care in the home, with necessary sevices provided, is as good as care in-hospital for acutely ill patients?”
Family physicians agreed that unscheduled telephone consultations (51.9%), distance to patients’ homes (67.1%), and home visits outside of office hours (53.2%) were barriers to providing acute care in the home. Additionally, 20.7% wrote in the “Other, please specify” section that poor remuneration was also a barrier.
Outcomes evaluation: comparing IPSITH and non-IPSITH groups
Controlling for the difference in symptom severity, there were statistically significant differences between the full set of IPSITH cases and non-IPSITH cases for ED visits both during (P = .002) and after (P = .034) the treatment period, as shown in Table 5, analysis 1. There were statistically significant differences between the subset of IPSITH cases and non-IPSITH cases for ED visits during the treatment period (P = .017) but not during the 6 weeks following (P = .092), as shown in Table 5, analysis 2.
Comparison of ED use between groups
After controlling for symptom severity, satisfaction with in-home services was significantly higher (P < .05) among the IPSITH group than the non-IPSITH group patients (91.0% vs 72.8%), caregivers (98.6% vs 87.5%), family physicians (98.7% vs 96.8%), and nurses (96.6 vs 83.6%), and there was no difference in the total caregiver burden score between the IPSITH group and non-IPSITH group (mean score out of 4 was 2.09 vs 2.12, P = .817).
Example IPSITH case
|
IPSITH—Integrating Physician Services in the Home.
Patients in both groups strongly preferred home care (80.2%) to hospital care (5.6%), but for family caregivers the preference was not as strong (73.1% vs 9.7%).
DISCUSSION
Inconvenient and expensive ED visits occurred less frequently for patients in the IPSITH program. On the one hand, IPSITH was associated with at least 14 fewer patient visits to the ED during the treatment period and 12 fewer ED patient visits during the 6 weeks following the treatment period. On the other hand, IPSITH used alternative resources, such as the 150 housecalls by the family physicians and the salary of the nurse practitioner. Broadly speaking this is the resource utilization trade-off.
Patients, caregivers, family physicians, and nurses were all more satisfied in the IPSITH group than in the non-IPSITH group. The IPSITH patients and family physicians expressed strong satisfaction with the nurse practitioner, whose role in IPSITH was pivotal.28 Despite the high satisfaction, IPSITH did not appear to reduce the burden felt by family caregivers.
Limitations
First, there was a need to impute severity for many non-IPSITH patients because of missing data; nonetheless another analysis was conducted on all non-missing data with similar results. Second, the generalizability of the findings might be limited in that the IPSITH physicians were not randomly allocated and were likely early adopters. In addition, the 29 early adopters subdivide into a group of 5 committed family physicians who admitted half of the 82 patients. We speculate physician characteristics such as older age (related to an elderly patient population), positive attitudes to home care, and lack of time and distance barriers contributed to this commitment.
Conclusion
This evaluation suggests that the new program of enhanced community-based home care that integrates family physicians and a multidisciplinary team is feasible, acceptable to all participants, and associated with positive outcomes. Future work needs to confirm this suggestion in larger samples and with criterion-standard quasi-experimental design or randomized control trials. As well, studies could further elucidate the burden for family caregivers and identify the barriers that prevented widespread involvement from family physicians in community-based health care programs. Initiatives aimed at overcoming barriers, particularly incentives for family physicians and confidence-building experiences for family physicians relative to their own clinical care and in-home services could be piloted and evaluated. Further research, then, should be aimed at evaluating the process of taking community-based health care programs to a wider group of family physicians and patients.
Based on the success of the IPSITH project, a similar program of acute care at home has been adopted by the Southwestern Ontario CCAC as the Advanced Home Care Team. In this model, nurse practitioners coordinate and deliver care in consultation with family physicians and other health care professionals.
Acknowledgments
Funding was provided by the Canadian Health Services Research Foundation, the Ontario Ministry of Health and Long-Term Care, the Community Care Access Centre of London and Middlesex, the Middlesex-London Health Unit, and the Thames Valley District Health Council. Dr Stewart is funded by the Dr Brian W. Gilbert Canada Research Chair in Primary Health Care.
Notes
EDITOR’S KEY POINTS
-
This study sought to examine a community-based model of care for seriously ill adult patients that incorporated patients’ family physicians. Patients with acute or complex illness received enhanced home-based care from a multidisciplinary team including their own family physicians in a novel care program called Integrating Physician Services in the Home (IPSITH).
-
Patient, caregiver, physician, and nurse participants were very satisfied with the program, but there was no difference in caregiver burden between the IPSITH and non-IPSITH groups, and family physicians cited unscheduled telephone consultations, travel, home visits outside of office hours, and poor remuneration as barriers.
-
Although IPSITH was associated with at least 14 fewer patient visits to the emergency department during the treatment period and 12 fewer emergency department visits during the 6 weeks following the treatment period, IPSITH required alternative resources, such as 150 housecalls made by the family physicians and the salary of a nurse practitioner.
POINTS DE REPÈRE DU RÉDACTEUR
-
Cette étude voulait examiner un modèle de soins intra-communautaires pour des patients adultes gravement malades avec la participation du médecin de famille du patient. Le patient souffrant de maladie aiguë ou complexe recevait de meilleurs soins à domicile d’une équipe multidisciplinaire comprenant son propre médecin de famille dans le cadre d’un nouveau programme appelé Integrating Physicians Services in the Home (IPSITH).
-
Les patients, médecins, soignants et infirmières participants étaient très satisfaits du programme, mais il n’y avait pas de différence dans la tâche de travail des soignants entre les groupes IPSITH et non IPSITH, et les médecins de famille ont indiqué que les consultations téléphoniques non prévues, les déplacements, les visites à domicile en dehors des heures de bureau et la rémunération insuffisante faisaient problème.
-
Même si le programme IPSITH permettait une réduction d’au moins 14 visites de patients à l’urgence durant la période de traitement et de 12 visites à l’urgence au cours des 6 semaines suivantes, il exigeait des ressources additionnelles telles que 150 visites de médecin à domicile et le salaire d’une infirmière praticienne.
Footnotes
-
This article is eligible for Mainpro-M1 credits. To earn credits, go to www.cfp.ca and click on the Mainpro link.
-
This article has been peer reviewed.
-
Cet article a fait l’objet d’une révision par des pairs.
-
Contributors
All authors were members of the research (evaluation) committee and were involved in the planning, execution, and evaluation of the project. All authors contributed to and approved the final manuscript. Dr Stewart was the principal investigator, responsible for all aspects of the research portion of the project. Dr Sangster was the Medical Director of the Community Care Access Centre and wrote the original proposal. Dr Ryan was the Research Project Coordinator for the project and conducted the analysis and participated in drafting the manuscript. Dr Hoch contributed to the analysis, in particular to the imputation of severity, and helped with the initial draft of the manuscript. Dr Cohen was the Medical Director for Integrating Physician Services in the Home (IPSITH), advised on the relevance of analyses, and created user-friendly presentations of results. Dr McWilliam was the lead researcher for the qualitative evaluation and advised on all aspects of the evaluation. Ms Mitchell was the nurse practitioner for the IPSITH project, advised on the relevance of analyses, and created user-friendly presentations of results. Dr Vingilis was responsible for the program evaluation framework and contributed to the analysis plan. Ms Tyrrell was the initial Program Coordinator for the project and was responsible for establishing the procedures for data collection. Dr McWhinney was responsible for the initial concept of the IPSITH program.
-
Competing interests
None declared
- Copyright© the College of Family Physicians of Canada