The prevalence of hyperglycaemia and its relationship with mortality, readmissions and length of stay in an older acute surgical population: a multicentre study.

BACKGROUND: The purpose of the study is to examine the prevalence of hyperglycaemia in an older acute surgical population and its effect on clinically relevant outcomes in this setting. METHODS: Using Older Persons Surgical Outcomes Collaboration (OPSOC) multicentre audit data 2014, we examined the prevalence of admission hyperglycaemia, and its effect on 30-day and 90-day mortality, readmission within 30 days and length of acute hospital stay using logistic regression models in consecutive patients, ≥65 years, admitted to five acute surgical units in the UK hospitals in England, Scotland and Wales. Patients were categorised in three groups based on their admission random blood glucose: <7.1, between 7.1 and 11.1 and ≥11.1 mmol/L. RESULTS: A total of 411 patients (77.25±8.14 years) admitted during May and June 2014 were studied. Only 293 patients (71.3%) had glucose levels recorded on admission. The number (%) of patients with a blood glucose <7.1, 7.1-11.1 and ≥11.1 mmol/L were 171 (58.4), 99 (33.8) and 23 (7.8), respectively. On univariate analysis, admission hyperglycaemia was not predictive of any of the outcomes investigated. Although the characteristics of those with no glucose level were not different from the included sample, 30-day mortality was significantly higher in those who had not had their admission glucose level checked (10.2% vs 2.7%), suggesting a potential type II error. CONCLUSION: Despite current guidelines, nearly a third of older people with surgical diagnoses did not have their glucose checked on admission highlighting the challenges in prognostication and evaluation research to improve care of older frail surgical patients.

Routine pharmacological venous thromboembolism prophylaxis in frail older hospitalised patients: where is the evidence?

It has been claimed that there are over 25,000 preventable in-hospital deaths from venous thromboembolism annually in the UK. NICE and SIGN guidelines therefore recommend that all hospitalised patients are risk assessed for venous thromboembolism. The guidelines would recommend using pharmacological thromboprophylaxis for all patients aged 60 and above with reduced mobility and acute medical illness unless obvious contra-indications exist. Meta-analysis data regarding pharmacological thromboprophylaxis for medical patients demonstrate reductions in asymptomatic deep vein thrombosis (DVT) rather than fatal pulmonary embolism and mortality. There is also the potential for increased bleeding risk with this approach. Evidence for older medical in-patients, particularly those aged over 75, is more limited being derived from subgroup analyses of larger clinical trials. In addition, based on exclusion criteria such as increased bleeding risk, frailer older adults were unlikely to have been included within such trials. This commentary will (i) critically appraise available data on the incidence of DVT and PE in older hospitalised patients; (ii) review the evidence available from meta-analyses and subgroup analyses in older medical in-patients for the use of venous thromboembolism prophylaxis; (iii) discuss those situations out-with the guidelines where venous thromboprophylaxis may not be appropriate and even potentially harmful in this patient group and (iv) suggest future research directions.