Table 1 Previous studies of body temperature and outcome after stroke - methods
First author and year | Study design | N | Types of stroke | Temperature measurement method | Time of 1st; interval; last reading | Statistical analysis | Outcome measures | Definition of pyrexia |
|---|---|---|---|---|---|---|---|---|
Azzimondi 1995[2] | P | 183 | Any stroke <48 hrs (not SAH) | A | N/S; 12 h; 7 d | Max temp in 7 d, logistic regression | CNS, serial GCS up to 30 d, 1, 3 & 6 month Barthel | ≤37.2 = absence of fever. ≥37.9 = high fever |
Reith 1996[3] | P | 390 | Any stroke <6 hrs | T | <6 h; -; - | Logistic regression | Lesion size, SSS, presence of infection, WCC | >37.5 |
Castillo 1998[9] | P | 297 | Ischaemic | A | <24 h; 2 h; 72 h | Correlate peak temp with clinical outcome and final infarct vol. Stepwise logistic regression | CSS, presence of infection, 4–7 d lesion volume, 3 month Barthel | >37.5 |
Georgilis 1999[4] | R | 330 | Any stroke | Most A, some RC | <48 h; 3 h; - | Presence/absence of fever and infection. Stepwise logistic regression | GCS, SSS, CT lesion volume, presence of infection, use of invasive procedures | >37.5 on >2 occasions on 2 consec. days |
Wang 2000[12] | R | 437 | Any “acute” stroke | T | “Admission” but no time limit given; -; - | Logistic regression. ischaemic vs. haemorrhagic stroke | Co-morbidities, WCC, [glucose], mortality (in-hospital and 1 yr) | >37.5 |
Boysen 2001[8] | P | 584 | Ischaemic | T | <6 h; 2–4 h; 48 h | Mean temp analysis by subgroup | SSS, 3 month mRS | >37.5 |
Kammersgaard 2002[6] | P | 390 | Any stroke | T | <6 h; -; - | Dichotomised normothermia vs. pyrexia, and multivariate survival analysis | SSS, [glucose], 5 yr mortality | >37 |
Audebert 2004[13] | R | 346 | Ischaemic; Excl pts with infection pre- or post- stroke | O or RC | <24 h (mean 6.7 h); 2–12 h; 3 d | Subgroup analysis of median temp days 1-3 | NIHSS, WCC, CRP, 1–5 d lesion volume (CT/MR) | ≥37.5 |
Sulter 2004[1] | P | 132 | Ischaemic; Excl pts on antibiotics on admission | RC | <12 h; continuous; 48 h | Dichotomised according to hyperthermia or not within 48 h | baseline NIHSS, presence of infection, effect of antipyretics | >37.5 |
Ernon 2006[14] | R | 107 | Ischaemic, with thrombo-lysis | T/O | <180 mins; random; 24 h; | AUC relative to 37° and to baseline T° | NIHSS, 3 month mRS | >37.4 |
Leira 2006[11] | P | 229 | First ischaemic; Excl pts with inflammatory or infectious disease | A | <24 h; -; - | Presence vs. absence baseline pyrexia | CSS, BP, blood biochemistry, 4–7 d lesion volume (CT) | ≥37.5 |
Wong 2007[15] | P | 156 | Ischaemic | T | <48 h (median 2.5 h); 4 h; 48 h | Mixed model, Lowess curves | Baseline NIHSS, use of paracetamol, presence of infection | None |
Idicula 2008[16] | P | 127 | Ischaemic with thrombo-lysis | N/S | <3 h; random; 5 d | Pre-thombolysis temperature and peak temperature in 5 d | NIHSS, 3 month mRS, BP, peak [glucose] | >37.7 |
Millan 2008[17] | R | 254 | Ischaemic stroke with thrombolysis | N/S | <3 h; 6 h; 48 h | Pre-thrombolysis, temperature at 24 and 48 h, and peak temperature within 24 h post- thrombolysis | NIHSS, 3 month mRS, early lesion vol (CT), MCA TIBI score (TCD) | ≥37 |
Saini 2009[7] | R | 5305 | First ischaemic | Most patients: A; others N/S | N/S; 8 then 24 hrly; 7 d | Normotherm vs. pyrexia at different time points | WCC, NIHSS, 3 month mRS, lesion vol (1–7 d CT/MR), use of antibiotics | >37.2 |
P RCT | 1399 or 1332 | Any stroke <12 hrs | T or RC | median 6 h, all <12 h; -; - | Multiple logistic regression | baseline NIHSS, 14 d Barthel, 3 month mRS | None | |
Naess 2010[20] | R | 250 (111 vs. 139) | Ischaemic | N/S | < 6 h; -; - | Logistic regression, temperature against outcome in tPA-treated vs. non-treated patients | baseline NIHSS, mRS on day 7 or at discharge, vascular risk factors, stroke aetiology | None |
Phipps 2011[21] | R | 1361 | Ischaemic, NIHSS ≥ 2 | Majority – T, partially unknown | <48 h; -; - | Logistic regression | vascular risk factors, NIHSS, stroke aetiology | ≥37.8 |