General Practice

Improving vaccine storage in general practice refrigerators

BMJ 1996; 312 doi: (Published 29 June 1996) Cite this as: BMJ 1996;312:1651
  1. Andrew Jeremijenko, general practitionera,
  2. Heath Kelly,
  3. Beverly Sibthorpe, epidemiologistb,
  4. Robyn Attewell, statisticianb
  1. a Department of General Practice, University of Western Australia, Claremont, WA 6010, Australia
  2. b National Centre for Epidemiology and Population Health, Australian National University, Canberra, ACT 0200
  3. Public Health Unit, Albany, WA 6330 Health Kelly, director.
  1. Correspondence to: Dr Sibthorpe.
  • Accepted 4 March 1996

Vaccines are biological products and are susceptible to fluctuations in temperature. In many general practices vaccines are exposed to adverse temperatures.1 2 The aim of this study was to determine whether educating one staff member in each practice about correct vaccine storage conditions and nominating that staff member to monitor the refrigerator's temperature would improve vaccine storage in general practices.

Methods and results

A random sample of general practices in a metropolitan division of Western Australia and all general practices in a rural division were invited to participate in the study. Those that already monitored their refrigerators with a maximum-minimum thermometer or did not store vaccines were excluded.

Practices were randomised into control and intervention groups, and their vaccine refrigerators were monitored for 30 days with a computerised temperature logger. The staff member most responsible for vaccine storage in each practice was then interviewed. In the intervention practices this staff member was educated, given a digital maximum-minimum thermometer, and allowed up to 14 days to adjust refrigerator temperatures if necessary. All practice refrigerators were subsequently remonitored with the computerised temperature loggers for a further 30 days. Monitoring occurred between October 1994 and January 1995.

A refrigerator was designated unacceptable if, during a 30 day monitoring period, more than eight days of logger readings were above 12°C or more than one hour of readings were below -0.5°C. These criteria were chosen to reflect exposure levels that may compromise the potency of vaccines.2 3 4

McNemar's test was used to assess whether refrigerator temperatures changed from baseline. Logistic regression with random effects5 was used to model the probability of a refrigerator being acceptable, accounting for group status (control or intervention) and repeated measurements (baseline and follow up).

Of the 36 metropolitan practices randomly selected, 29 were eligible and 25 agreed to participate. Of the 28 rural practices identified, all 25 eligible practices agreed to participate, giving a total of 50 practices. Only five of the staff members responsible for vaccine storage in these 50 practices knew the recommended maximum and minimum temperatures for vaccine storage and only two knew which vaccines were damaged by freezing.

Changes in the acceptability of refrigerator temperatures at baseline and follow up are shown in table 1. Logistic regression analysis showed that the odds ratio of a refrigerator in the intervention group being acceptable at follow up relative to baseline was 6.8 (95% confidence interval 1.9 to 24.3), while in the control group the odds ratio was 0.6 (0.2 to 1.7). The interaction between group status and acceptable temperature at follow up was statistically significant (χ12 = 8.4; P = 0.004). Of the 30 unacceptable refrigerator recordings at baseline and follow up, 26 recorded more than one hour below -0.5°C.

Table 1

Number of vaccine refrigerators with unacceptable temperatures at baseline and follow up in general practices in Western Australia, October 1994 to January 1995

View this table:


Eighteen (36%) practice refrigerators recorded unacceptable temperatures at baseline: most of these refrigerators recorded temperatures that may have frozen vaccines and consequently damaged them. The staff members responsible for vaccine storage had poor knowledge about recommended temperatures and were unaware of their refrigerator's temperatures. This randomised controlled trial shows that educating a staff member in each practice on correct vaccine storage conditions and nominating that staff member to monitor the refrigerator's temperature with a digital maximum-minimum thermometer improved vaccine storage. Selection bias was minimised as only four eligible practices refused to participate. Blinding was considered unnecessary as recordings were by a computerised device, and though only a small sample was studied a significant improvement was nevertheless observed. The widespread implementation of this simple and inexpensive intervention should result in better vaccine storage conditions and fewer vaccine failures in general practices.


  • Funding Lederle, Commonwealth Serum Laboratories, and Hastings Data Loggers.

  • Conflict of interest None.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
View Abstract