• Open Access

Acute hospital admission for sepsis: an important but under-utilised opportunity for smoking cessation interventions


Correspondence to:
Dr Joshua S. Davis, John Mathews building (Building 58), Royal Darwin Hospital, Rocklands Drive, Casuarina, NT 0810, E-mail: Joshua.davis@menzies.edu.au

Although an integrated health promotion approach is the keystone of effective tobacco control, acute care hospitals do have a specific role to play in the individual screening, risk assessment, and education of patients about smoking cessation.1–2 A systematic review of hospital-based interventions estimated that counselling about smoking cessation commenced during hospitalisation and continued for at least one month following discharge increases the odds by 65% that person remains a non-smoker at 6–12 months.2 Clearly, the first step in any hospital-based intervention aimed at smoking cessation is to identify smokers. Asking patients about tobacco smoking is considered a routine part of medical history taking, as is documentation of this information in the patient's medical record.

In 2007/08 we performed a 12-month prospective epidemiological study of community-acquired sepsis at Royal Darwin Hospital. Demographic and clinical information was prospectively collected from medical records and hospital databases. All adult patients admitted with a suspected or proven infection along with two or more criteria for the systemic inflammatory response syndrome (SIRS)3 were enrolled. A total of 1,191 admissions were included in the study, of whom 604 (50.7%) were Indigenous and 624 (52.4%) were male. Of these 1,191 acute hospital admissions, 391 (31.9%) had no documentation of smoking status. Of the remaining 800, 413 patients (51.6%) were documented as current smokers, 380 (47.5%) as non-smokers, and seven had missing data due to lack of access to the patient's medical record. This smoking prevalence of 51.6% is substantially higher than the NT age-adjusted prevalence of 35.3% (p<0.001),4 suggesting that smoking may be a risk factor for sepsis-related hospitalisation. Smoking was more common in males (63.2%) than females (51.2%) and in Indigenous (66.5%) than non-Indigenous (49.1%) patients (Table 1).

Table 1.  Crude prevalence of smoking by Indigenous status and sex.
Crude prevalence%MaleFemaleTotal
Indigenous71.4 (n=130/182)62.4 (n=136/218)66.5
Non-Indigenous57.4 (n=147/256)33.6 (n=46/137)49.1
Combined63.2 (n=277/438)51.2 (n=182/355)52.0

We also examined predictors of a lack of documentation of smoking status. Smoking status was documented in 60.1% of those admitted under a surgical team compared with 74.3% of patients under a medical team (p<0.001). In a multivariate model controlling for comorbidities, age, gender, severity of sepsis and homelessness the admitting team remained a statistically significant predictor of a lack of documentation of smoking status (odds ratio of 2.30 [95% CI 1.71–3.08] for surgery versus medicine). Documentation of smoking status was also less likely (p<0.05) in females, those reporting hazardous alcohol use, and those with dementia, but not in those with more severe sepsis. Being admitted under a surgical team and hazardous alcohol use were the only significant predictors of a lack of documented smoking status in our multivariate model. A potential limitation of this data is that a lack of documentation does not prove that smoking status was not discussed. However, it is more likely that the admitting team or alcohol use are associated with failure to discuss smoking status rather than simply a failure to document such a discussion.

Acute care smoking assessment and intervention is only one part of the overall integrated approach to tobacco control. However, the potential benefits are likely to be significant, with tobacco control being an extremely cost effective intervention.1 There are many reasons that smoking cessation interventions are overlooked in acute hospital admissions such as lack of time or resources, and poor systems. The alarmingly high rates of smoking in this cohort, particularly among Indigenous people, strengthens the case for improving acute care smoking assessment and cessation management, with a particular focus on improving the assessment of surgical patients. For many people, especially the most vulnerable, hospitals are often the first if not the only point of contact they have with the health system. It is time that acute care hospitals stop wasting this important opportunity.