Current issues of ACP Journal Club are published in Annals of Internal Medicine


Vaccination of health care workers reduced geriatric patient mortality

ACP J Club. 1997 Jul-Aug;127:1. doi:10.7326/ACPJC-1997-127-1-001

Source Citation

Potter J, Stott DJ, Roberts MA, et al. Influenza vaccination of health care workers in long-term-care hospitals reduces the mortality of elderly patients. J Infect Dis. 1997 Jan;175:1-6.



To determine whether vaccination of health care workers (HCWs) employed in geriatric medical long-term care hospitals and vaccination of long-term patients reduces incidence of influenza, lower respiratory tract infection, and death in patients.


Randomized controlled trial (factorial design) with 5-month follow-up.


12 geriatric medical long-term-care hospitals in Glasgow, Scotland.


1059 patients (mean age 77 y, 71% women) who were residents in a geriatric medical hospital.


6 hospitals were allocated to a vaccination program where HCWs were vaccinated, and 6 hospitals were allocated to a no-vaccination program. 1078 HCWs were involved in the study. Hospital policies of vaccinating (6 hospitals) or not vaccinating patients (6 hospitals) were maintained according to the usual practice.

Main outcome measures

Patient mortality and occurrence of influenza-like illness.

Main results

In hospitals where HCWs were offered vaccination, 653 (61%) were vaccinated. Only 1 patient was vaccinated in a hospital with a policy of not routinely vaccinating patients compared with 478 (89%) in hospitals where patients were routinely vaccinated. Fewer patients died in hospitals where HCWs were vaccinated than in hospitals where HCWs were not vaccinated (P < 0.001) (Table), and fewer patients developed suspected influenza-like illness { P < 0.001}*. No difference was shown in patient mortality in hospitals where patients were offered vaccine and those where patients were not offered vaccine (P = 0.36). In a multivariate analysis, nonvaccination of HCWs (P < 0.001), a lower Barthel score (P = 0.004), and increased age (P = 0.03) of the patient were associated with increased mortality; hospital policy for nonvaccination of patients was not associated with increased mortality (P = 0.33).


Vaccination of HCWs employed in geriatric long-term care facilities was associated with a reduced rate of patient mortality and influenza-like illness.

Sources of funding: Greater Glasgow Health Board Care of the Elderly Unit and Medeva.

For article reprint: Professor D.J. Stott, Academic Section of Geriatric Medicine, Royal Infirmary, Glasgow, G4OSF, Scotland, UK. FAX 44-141-211-4944.

* P value calculated from data in article.

Table. Vaccinated health care workers (HCWs) vs nonvaccinated HCWs†

Outcome Vaccinated HCWs EER Nonvaccinated HCWs CER RRR (95% CI) ARR |EER-CER|
Patient mortality 10% 17% 41% (19 to 57) 7%

†Abbreviation defined in Glossary; RRR, ARR, and CI calculated from data in article.


If you think the epidemiology of influenza has been fully described, think again. Potter and colleagues have provided some very interesting and important new findings. The trial by Potter and colleagues appropriately randomized geriatric long-term-care facilities rather than the individual patients. Vaccination of some HCWs and not others in the same institution would have resulted in partial herd immunity and confounded the determination of communicability. This study was not designed to show vaccine efficacy in the patients themselves and did not have the statistical power to do so. Understanding the components of the Barthel score and how it was weighted would have permitted the reader to better assess the mortality findings.

The process of vaccinating HCWs made a substantial difference in reducing mortality among the patients. When patients were not vaccinated, vaccination of HCWs resulted in fewer patient deaths from all causes. Even when patients were vaccinated, hospitals with vaccinated HCWs had fewer patient deaths than hospitals with unvaccinated HCWs.

A positive vaccine effect was also observed when the authors studied the development of influenza-like illnesses in the patients. In hospitals where HCWs were vaccinated, influenza-like illnesses occurred in 7.7% of unvaccinated patients compared with 0.9% of vaccinated patients. When the HCWs were not vaccinated, no difference was observed between the vaccinated and unvaccinated patients (6.2% vs 8.8%); why this occurred is not clear. Numerous other studies show a positive effect of vaccination in infirm elderly patients (1-3).

The take-home message is that HCWs who come in contact with infirm elderly persons should be immunized with the influenza vaccine.

Peter A. Gross, MD
Hackensack University Medical CenterHackensack, New Jersey, USA


1. Fedson DS, Wajda A, Nicol JP, et al. JAMA. 1993;270:1956-61.

2. Final results: Medicare influenza vaccine demonstration—selected states, 1988-1992. MMWR Morb Mortal Wkly Rep. 1993; 42:601-4.

3. Gross PA, Hermogenes AW, Sacks HS, Lau J, Levandowski RA. Ann Intern Med. 1995; 123:518-27.