Hypothermia in elderly patients presenting to Accident & Emergency during the onset of winter

SMJ 2002;47(1): 10-11

Abstract: It is well established that the elderly population is vulnerable to hypothermia, leading to increased morbidity. A prospective observational study took place between 1 October 1999 and 31 December 1999 in a large teaching hospital Accident and Emergency department. Core temperature was assessed at presentation using a tympanic probe on patients over 65 years of age. A total of 1543 eligible patients attended during the study period. Complete data was available on 958 patients. Forty-eight patients (5%) were found to be hypothermic (core temperature <350Celcius). There were two peaks in hypothermic presentations; these corresponded to periods of cold weather. The incidence of hypothermia was higher in non-ambulant patients and those with co-morbidity; the majority of patients lived in relatively deprived areas by a postcode derived deprivation index. Mortality was 34% in patients hypothermic at presentation.

Hypothermia is recognized to have a range of adverse physiological effects which may contribute to excess mortality in the winter months.1 Physiological studies have shown the elderly to be particularly vulnerable to the effects of hypothermia.2 Many elderly patients are unable to adequately heat their homes.3 Although uncertainty over winter fuel allowance has recently been ameliorated with the introduction of annual winter fuel payments to all households of those over 65 years, financial insecurity remains an important issue in the elderly.4

A previous study into the epidemiology of hypothermia in Scotland identified the elderly patient as being at particular risk.5 Hypothermia may be a factor in acute decompensation, and so lead directly to hospital admission, or may be an incidental finding in patients with unrelated pathology.

A prospective observational study was carried out of tympanic membrane temperature in patients over the age of 65 years attending the Accident and Emergency department. Tympanic membrane temperature was measured by the triage nurse at presentation using the Genius IItm probe, and was recorded along with the maximum and minimum daily ambient temperature during the study period of 1 October to 31 December 1999.

Four Genius IItm probes were calibrated against a bench standard at the beginning of the study, and nursing staff trained in the use of the probes, according to the manufacturer’s guidelines.

For each hypothermic patient, data regarding presenting complaint, mobility, home circumstances, and post-code was collected by questioning the patient and their family and referring to Scottish Ambulance Service records. Subsequently, patients were followed up to discharge from hospital or death. A measure of the patients’ socioeconomic status was derived from the post-code using the Carstairs index.6

A total of 1543 eligible patients attended during the study period. Complete data was available on 958 as detailed in Table I. Forty-eight patients (5%) were found to be hypothermic (core temp <35⁰ Celsius) at presentation. Although hypothermic patients presented throughout the study period, the rate of presentation varied according to the ambient temperature. Two peaks were evident. The first corresponded to the early frosts of October the second to a period of sustained cold in mid December. (See Figure 1.)

Table I  

Of the forty-eight patients presenting with temperatures less than 350 Celsius, thirty-eight required acute admission to an inpatient specialty. Sixteen patients (34%) died. Of those who survived the median stay till discharge was 4 days (range 1 - 20days) as shown in Table II.

Table II

Only four patients were found outside, the remainder being admitted from their homes. Twenty-six patients lived alone. Five patients were admitted from residential or sheltered accommodation. Twenty-eight patients were known to have limited mobility or were described as house bound prior to presentation. A postcode derived deprivation measure (DEPCAT)6 (Table III) was used to categorize the living conditions of the forty-eight patients. Seventy-one per cent of hypothermic patients were found to be in the relatively deprived categories 3 – 6 although this comprises only 44% of the housing in Tayside. (See Table III.)

Table III

Previous studies have attempted to define the incidence of hypothermia in the Scottish population. Hislop et al suggested a presentation rate of 1 per 14,000 people per winter, based on analysis of the presenting complaints of patients attending Accident and Emergency departments in the West of Scotland.5  In our study of 1543 patients over the age of 65 years attending A&E with a range of problems, tympanic temperature data was collected on 958 patients, representing only 65% of eligible patients. There may have been a tendency to omit temperature recording in patients presenting with minor injuries, resulting in an overestimate of the proportion of patients presenting with hypothermia. Indeed, 5% of elderly patients in our study were found to be hypothermic at presentation, a much greater prevalence than suggested by Hislop et al.

Tympanic temperature measurement has been shown to be superior to rectal temperature in reflecting core temperature,7 and allows a rapid, non-invasive, digital assessment. This method of temperature measurement is likely to prove more acceptable than rectal recording for screening for hypothermia.

Excess mortality in the elderly population during the winter months has been related to environmental conditions. Observed excess mortality may be due in part to acute coronary syndromes and stroke, postulated mechanisms to account for this include a seasonal rise in mean arterial blood pressure.1 This ‘cold stress’ may affect patients with advanced cardiovascular morbidity resulting in decompensation. There may also be a contribution to peaks in hospital admission; 79% of patients hypothermic at presentation required acute admission compared to the overall admission rate of fifty percent of patients over 65 in our department. Alternatively, it may be that a temperature of below 350C is a marker of illness severity, and is, therefore, an incidental finding in patients with other medical problems. Mortality in hypothermic patients was high (34%).

The majority of patients in our study were admitted from their own homes. A large proportion of patients lived alone. This is unsurprising since socially isolated elderly patients with various medical problems and associated poor mobility are noted to be at greatest risk of hypothermia3.

There continues to be debate regarding fuel poverty in the elderly population. Previous studies have shown fuel cost to be a considerable worry to the elderly, and this may lead to inadequate domestic heating.4 Seventy-one percent of the population found to be hypothermic at presentation were from relatively deprived postcode regions. It may be inferred from recent population studies that better preparation for cold weather in the form of domestic heating has an impact on mortality.8 Recent proposals by the Scottish Executive to provide central heating for all elderly people are a welcome initiative.

References

Back to February Index