I read with interest the article by Skinner et al1 assessing whether case management of frequent attenders to an urban emergency department reduces frequency of subsequent attendances. The case for multidisciplinary management of frequent attenders is well made, and the reported practice of case review and implementation of care plans is highly commendable.

However, a possible considerable limitation of this study is not fully elucidated in the authors’ discussion. Although mention is made that the reduction seen in attendances by frequent attenders after the case review “may be due to the natural ebb and flow in the presentations of these patients”,1 comment is not made on the potential substantial effects of the differing seasons in which the study was conducted (October 2006 to March 2007, control, compared with April 2007 to September 2007, postintervention).

Watters et al2 have previously observed that the number of patient attendances to the emergency department at the Royal Infirmary Edinburgh is subject to monthly variation. χ² analysis of the total numbers of patients presenting each month presented by Watters et al confirms the variation in numbers of patients by month (χ² = 105.24, p<0.001), even when the varying numbers of days in each month is taken into account (χ² = 57.03, p<0.001).

Further to this variation in total attendances is the seasonal variation in presentations of different clinical conditions. The most common documented diagnoses in this group of frequent attenders were stated to include “alcohol-related problems (46%), mental health problems (37%) and chronic complaints such as abdominal or chest pain (40%)”. Other diagnoses included chronic obstructive pulmonary disease/asthma and drug abuse. However, the presentations of many of these conditions have been shown to be subject to seasonal variation. The study of Arfken et al3 of frequent visitors to psychiatric emergency services found that admission was more frequent in inclement weather. Halpern et al4 observed a seasonal variation in emergency department “psychiatric visits”, with results supporting the “existence of a Christmas effect”; they also noted that substance misusers were more likely to attend the emergency department during the weeks surrounding Christmas.

It is also known that patients with ischaemic heart disease are more likely to become severely symptomatic during colder temperatures, and “cold stress” contributes to a higher mortality rate from this condition during the winter.5 Winter predominance seasonal variation in presentation rates for abdominal pain is also suspected and has been demonstrated in children.6 Ballester et al7 observed “clear temporal patterns of COPD emergency”, with more emergencies in winter.

With more than half of the frequent attenders’ documented diagnoses subject to seasonal variation with increased presentations during winter months, it would appear that there is an as yet unaddressed confounding bias in the observed decreased number of subsequent attendances in the (summer) months after case management intervention. I eagerly await the mentioned follow-up study and will be interested to ascertain whether the results seen so far are indeed sustainable over time.