| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
|
|
|
|||||||
|
|||||||||
a Division of Geriatric Medicine, Department of Medicine, University of Alberta, Edmonton, Canada
b Glennan Center for Geriatrics and Gerontology, Eastern Virginia Medical School, Norfolk
Janet E. McElhaney, Glennan Center for Geriatrics and Gerontology, Eastern Virginia Medical School, Hofheimer Hall, Suite 201, 825 Fairfax Avenue, Norfolk, VA E-mail: mcelhaje{at}evms.edu.
William B. Ershler, MD
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
Methods. An influenza vaccination program using a standing order policy was implemented to vaccinate all consenting persons 65 years and older prior to hospital discharge. This was a prospective, single center, cohort study in a tertiary care university teaching hospital during an 8-week vaccination period in the fall of 1994 and follow-up during the subsequent influenza season. The vaccination status of each patient was recorded as no vaccination, vaccination received in hospital, or vaccination in the community prior to or after the hospitalization. Hospital vaccination rates were compared with the rate of vaccination of older adults in the community. During the subsequent influenza season, the number of subjects reporting symptoms of influenza-like illness (ILI) or who were readmitted to hospital with influenza-related illness was compared in an analysis of vaccinated versus unvaccinated subjects.
Results. Seven hundred and sixty-one patients were interviewed, and 332 of these individuals had been vaccinated in the community prior to their hospital admission. Of the remaining 429 unvaccinated patients who were eligible for vaccination in the study, 171 were vaccinated in our immunization program, eight were vaccinated in the community after discharge, and 244 were not vaccinated. We were able to increase the absolute vaccination rate by 22%, when compared with community rates, with our immunization program. The number of subjects with ILI symptoms or readmission to hospital was too small to compare the vaccinated to the unvaccinated group in the study.
Conclusions. An inpatient influenza immunization program using a standing order policy was able to target a particularly high-risk subset of persons 65 years and over who might otherwise have not received influenza vaccination.
INFLUENZA and related illnesses continue to be an important cause of morbidity and mortality in the over-65 population (1). Previous studies have documented that influenza vaccination of older adults reduced the number of hospitalizations for pneumonia and respiratory illnesses, influenza-related illnesses, and all-cause mortality (2)(3). Furthermore, it has been shown that there is a reduction in the complications of influenza in all vaccinated persons 65 years and older, not only in those at greatest risk (4). Although the Advisory Committee on Immunization Practices and the National Advisory Committee on Immunization (Canada) recommend annual influenza vaccinations for persons 65 years and over and for persons with conditions placing them at high risk for complications of influenza, less than 40% of persons 65 years and older were vaccinated in Canada and the United States each year (5).
In 1987, Fedson suggested that to improve immunization rates, educational programs, administrational and organizational changes and various incentives and disincentives were required (6). Subsequent studies have elucidated some of the factors associated with receipt of vaccination: being 70 years old or older, possessing the targeted medical conditions for influenza vaccination, and having visited a physician in the previous year (7). Strategies to alter physician practices and recommendations from other health care providers to promote influenza vaccination in high-risk groups have been shown to improve vaccination rates (8)(9). Particularly in persons with high-risk chronic medical conditions, vaccination has been shown to be cost-effective (10)(11). However, it has been shown that older persons who are hospitalized during the period when influenza vaccination is being administered in the community have low rates of vaccination and a high risk of readmission to the hospital during the subsequent influenza season (12). Thus, the purpose of this study was to institute an influenza vaccination program that would increase vaccination rates above those in the community in persons aged 65 and over who were hospitalized during the period of vaccination in the fall of 1994.
A review of our hospital's data on influenza vaccination rates among hospitalized older adults showed that in-hospital vaccination rates were well below 5%. The main barrier to vaccination appeared to be the requirement for a physician order; most of the medical staff did not view vaccination as a priority or were concerned that vaccination might not be effective or might complicate the patient's course in hospital. Because an educational program was predicted to be ineffective for changing in-hospital practices of the attending staff in our tertiary care center, we designed an influenza immunization program using a standing order under the principal investigator for the study (JM). The purpose of the study was to increase vaccination rates in this very high-risk group of hospitalized older adults. In addition, to determine the benefit of vaccination, we compared vaccinated to unvaccinated subjects in the study for the incidence of influenza-like symptoms, influenza-related illnesses, hospitalizations, and death during the influenza season in our community, which started approximately 8 weeks after the end of the vaccination program.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
All patients visited by the study personnel agreed to participate in the study, whether they were vaccinated in hospital or not. A standardized questionnaire was used to interview all subjects including those vaccinated before admission and those refusing immunization. The reasons for the current hospitalization and past medical history were obtained directly from each participant and from his or her medical records. Specific notes were made of significant reactions to eggs or egg proteins, medications, previous influenza vaccination, as well as a history of smoking, alcohol intake, diabetes mellitus, hypertension, heart disease (ischemia, valvular, dysrhythmia), pulmonary disease (chronic obstructive pulmonary disease, asthma, or other), kidney disease, malignancy, or stroke.
Vaccination
A standing order was used to give the vaccine because there had been previous failed attempts in this hospital to obtain physician orders for influenza vaccination. Also, this was the first time that a program had been instituted in which a treatment was given without a written order from the attending physician. If consent was given by the patient, vaccination with a whole virus preparation of the 1994–1995 influenza vaccine was administered intramuscularly using a standing order from the physician (JM) involved in the study. Thus, no order was required from the attending physician, although physicians not agreeable to this could have their patients withdrawn from the study. No physicians withdrew their patients from the study. Permission to vaccinate under a standing order was obtained from the medical staff advisory board before the study commenced. The 1994–1995 influenza vaccine used was a whole virus preparation containing 15 micrograms each of A/Shangdong/09/93, A/Texas/3691, and B/Panama/45/90 (Biovac Division of Biochem Pharma, Montreal, Canada). Patients who were vaccinated prior to or after admission to the hospital received a split virus vaccine containing the same viral antigens in the community; the date of vaccination could not be accurately recorded in all cases.
Follow-Up for Flu-Like Illness
Vaccine was administered once the patient had recovered from the acute phase of his or her illness and before discharge. The time of vaccination in hospital was recorded. A card was provided to each participant vaccinated in hospital recording the date of vaccination and a contact person (the research nurse) should any questions or problems arise pertaining to the immunization. All participants, regardless of vaccination status, were asked to notify the research nurse during the following winter months if they developed flu-like symptoms or an acute respiratory illness determined by the development of three or more symptoms including fever, cough, confusion, chills, headache, sore throat, and muscle and joint aches. In addition, readmissions of subjects to the study hospital during the influenza season were further assessed, and specimens were collected by the study personnel within 24 hours of admission. Patients were assessed for any influenza-like symptoms, and specimens for laboratory investigations necessary for the diagnosis of influenza illness were collected. Diagnostic tests included nasopharyngeal swabs for virus detection by a direct fluorescence assay (Provincial Laboratory, University of Alberta, Canada) and acute and convalescent serology to detect a fourfold rise in antibody titer to the circulating strain of influenza virus.
At the end of the influenza season (March 1995), follow-up telephone calls were made to all subjects. The number of patients reporting influenza-like symptoms during the influenza season was recorded. The number of deaths and hospital readmission rates for influenza-related illness during the influenza season was also obtained for this group by telephone contact with the patient or family physician. Influenza-related illness included influenza pneumonia, delirium, failure to thrive, exacerbation of COPD, and congestive heart failure.
Statistical Methods
Analyses of the data in this study were done using the EPI Info Statistical Package (EPI-Info 6, Centers for Disease Control, Atlanta, GA). Categorical variables were compared by the chi-square test and, as multiple comparisons were made, statistical significance was defined as p < .01.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
The risk factor profile for the vaccinated group was significantly different when compared with the unvaccinated group as shown in Table 1 . The vaccinated group contained a greater percentage of patients deemed at high risk of influenza due to such chronic illnesses as cardiac and respiratory conditions, diabetes mellitus, and renal dysfunction. There was no difference in risk factor profile for the group vaccinated in hospital with whole virus vaccine compared with the group who received split virus vaccine in the community (data not shown).
|
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
A subsequent study the following year comparing this hospital site with a similar hospital site without a standing order policy showed that although similar numbers of patients consented to receive the influenza vaccine, only one third of patients were vaccinated if an attending physician order was required. This result compared with an over 90% vaccination rate in consenting patients where a standing order policy for influenza vaccination was in place (J. McElhaney, unpublished data, 1995), thus demonstrating a significant barrier to vaccination when an individual physician order is required. These two sequential years of in-hospital vaccination programs provided the groundwork for a subsequent hospital-based pneumococcal vaccination program in which vaccination was automatically ordered under a standing order policy and given to all patients who had not previously received the vaccine. This program was much more cost-effective because written informed consent by study personnel was not required, and even higher vaccination rates in the range of 90% were achieved.
Generally, the vaccinated group had more chronic diseases than the unvaccinated group. It may be postulated that patients with medical conditions that place them at high risk of influenza-associated complications are more likely to have had a physician visit during the period influenza vaccination was promoted (7)(9). Because our study showed a similar case mix in the group vaccinated before their hospital admission compared with the group vaccinated in hospital, this suggests that there remains a barrier to vaccination of these high-risk individuals in the community. When presented with the risk/benefit profile of vaccination during a hospital admission, these individuals may be more willing to accept any potential side effects of vaccination.
Although some may argue that hospitalized patients may mount less of a response to the vaccine than patients convalescing in the community after discharge from hospital, previous studies have shown that the response to vaccination in hospitalized, convalescing patients is comparable to those vaccinated 1 month after discharge (17). Fedson has also demonstrated that hospital admission, in addition to age and underlying high-risk conditions, is another marker for identifying patients at greatest risk for influenza-related illnesses. During the influenza outbreak period, 39%–46% of patients admitted with influenza-related conditions had been discharged from the hospital at least once during the previous vaccination season (18). In fact, patients admitted to the hospital during the vaccination period have a 3.5% readmission rate for influenza-related illness and a 10% readmission rate for acute respiratory conditions (12). Thus, our intervention had the potential to prevent 17 hospital admissions for acute respiratory conditions. In addition, these types of vaccination programs present an opportunity for targeting a high-risk population for other preventive treatments such as pneumococcal vaccination.
After demonstrating that an organized immunization program can significantly improve vaccination rates, the next stage would be to demonstrate positive outcomes in terms of morbidity and mortality. However, because the 1994–1995 influenza season was relatively mild (19), there were too few readmissions to hospital to derive any meaningful interpretation of efficacy of vaccination on readmission rates and incidence of influenza-related illness. In addition, insufficient data were available on causes of death in the vaccinated and unvaccinated groups to determine the effect of vaccination on mortality rates. Death certificates were not available to us to review and assess primary and secondary causes of death. Without objective evidence of the presence of influenza in the form of increasing convalescent viral titer or positive nasopharyngeal cultures, it would have been difficult to directly attribute influenza as a contributor to the cause of death. However, review of the literature leaves little doubt about the benefits of influenza vaccination in at-risk older adults. A randomized, double blind, placebo-controlled trial of influenza vaccination in a community setting confirmed that vaccine recipients had a twofold decrease in risk of respiratory illness, with even greater protection in those who had chronic underlying medical conditions or a previous vaccination history (20). Nichol also provides convincing evidence of consistent reductions in all-cause mortality and hospitalizations for pneumonia, congestive heart failure, and all respiratory conditions over six influenza seasons (21).
The limitations of this study are related to inadequate follow-up data to document influenza-related illness and acute respiratory conditions and demonstrate the benefit of our vaccination program. There were several barriers to adequate follow-up. Geographically dispersed subjects (the hospital's referral base covered Northern Alberta) limited the possibility of population-based influenza surveillance by study personnel, and individual subjects were lost to follow-up. Thus, the actual benefit of our program must be projected from previous studies of vaccine efficacy.
In summary, a standing order policy for influenza vaccination can markedly improve the use of this proven, efficacious, preventive intervention in a vulnerable inpatient population. Future work comparing the vaccination rates and outcomes in hospitalized older adults randomized to an immunization program versus a "usual care" group may be required. However, our experiences at one large, tertiary care center suggest that optimal vaccination programs can achieve 90% vaccination rates in hospitalized older people. This is a much more practical and worthwhile endeavor in light of the overwhelming evidence for the benefits of influenza vaccination in this population.
Conclusion
An inpatient influenza immunization program was able to target a particularly high-risk subset of persons aged 65 years and over who might otherwise have not received influenza vaccination.
|
Acknowledgments |
|---|
Received February 19, 1998
Accepted October 22, 1999
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  D. C. Ompad, S. Galea, and D. Vlahov Distribution of Influenza Vaccine to High-Risk Groups Epidemiol. Rev., August 1, 2006; 28(1): 54 - 70. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Traeger, A. Thompson, E. Dickson, and A. Provencio Bridging Disparity: A Multidisciplinary Approach for Influenza Vaccination in an American Indian Community Am J Public Health, May 1, 2006; 96(5): 921 - 925. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. A. Winston, M. C. Lindley, and P. M. Wortley Lessons learned from inpatient vaccination in michigan. American Journal of Medical Quality, March 1, 2006; 21(2): 125 - 133. [Abstract] [PDF]
|
|
|
|
 |
|
 P. R. Dexter, S. M. Perkins, K. S. Maharry, K. Jones, and C. J. McDonald Inpatient Computer-Based Standing Orders vs Physician Reminders to Increase Influenza and Pneumococcal Vaccination Rates: A Randomized Trial JAMA, November 17, 2004; 292(19): 2366 - 2371. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| All GSA journals | The Gerontologist |
| Journals of Gerontology Series B: Psychological Sciences and Social Sciences | |
