P244 Risk stratification of flu in post-pandemic winter 2010

    loading  Checking for direct PDF access through Ovid



The aim of the study was to review the assessment of acute medical admissions with “Flu like illness” and to identify useful tools in risk-stratifying severity of illness.


This was a retrospective observational study. We reviewed the assessment of all inpatients diagnosed with “flu like illness” from November 2010 to March 2011 at a district general hospital. We evaluated potential risk-stratification tools with respect to adverse outcomes (length of admission and intensive care unit (ITU) admission): Co-morbidities (diabetes mellitus, immunosuppression, pregnancy, chronic respiratory, heart, renal and liver disease), CURB-65 score, C-reactive protein (CRP) and chest x-ray findings (CXR).


27 patients were identified; 6 Male, 19 Female; mean age 40.1. 22 had virology swabs; 12 (62%) were positive for H1N1, 5 (24%) Influenza B and 2 (9%) Influenza A. Length of admission (LOA) ranged from 1 to 30 days (mean 8.3 days). Seven patients required ITU admission. 14 (52%) had no co-morbidities. Six (86%) of seven ITU patients had no co-morbidities. LOA did not differ between patients with co-morbidities and those without (10 vs 10.9 days respectively). 17 (63%) patients had CURB-65 of zero. Five (71%) of seven ITU patients had a CURB-65 of less than three. CURB-65 was poorly correlated with LOA (R2=0.22). CRP on admission ranged from <4 to 511 mmol/l (mean 121 mmol/l). Mean CRP of ITU patients was 240 mmol/l; in contrast to 79 mmol/l in non-ITU cases. CRP was poorly correlated with increased LOA (R2=0.16). 25 patients had CXR on admission and 12 (48%) had abnormal findings. Patients with bilateral CXR changes had a mean LOA of 21 days compared with 3.5 days in those with normal CXR. Six (86%) of seven ITU patients had abnormal CXR.


A raised CRP and abnormal CXR findings on admission were associated with adverse outcomes. Co-morbidities and CURB-65 correlated poorly with disease severity. These findings may be explained by the high prevalence of H1N1 influenza in winter 2010. Current Health Protection Agency guidelines place strong emphasis on CURB-65 and co-morbidities in risk-stratification. We recommend the inclusion of CRP on initial assessment and stronger emphasis on CXR changes.

Related Topics

    loading  Loading Related Articles