Day case surgery is increasing in modern healthcare practice, accounting for an estimated 75% of all surgical procedures in the USA in 1995 . Quality clinical care and cost efficacy require that the incidence rate of perioperative morbidity is minimized. In the perioperative period, headache is an important cause of minor morbidity with an incidence rate of up to 70% . Studies of perioperative headache to date may not reflect the current organization of the day surgery unit and modern anaesthetic techniques. Our objective was to identify qualitative risk factors for perioperative headache in patients undergoing day case surgery.
This study was approved by our Hospital Ethics Committee. We invited 230 consecutive patients attending for day case surgery in our purpose-built day surgery unit to complete a questionnaire about their previous experience of headache and certain aspects of their lifestyle. Patients started filling in the questionnaire upon arrival and were asked (by their anaesthetist or day unit nurse as appropriate) whether they had a headache on four separate occasions: (1) on arrival; (2) before induction of anaesthesia; (3) in the recovery room on emerging from anaesthesia; or (4) before discharge.
The surgical case-mix included gynaecology, urology, orthopaedics and general surgery. Caffeine consumption (mg day−1) was calculated based on the following conversion factors: fresh coffee, 120 mg caffeine cup−1; instant coffee 90 mg cup−; tea 50 mg cup−. Alcohol consumption was also assessed. One unit of alcohol was taken to be equivalent to half a pint (284 mL) of average-strength beer, a single measure (25 mL) of spirits or one glass of wine (i.e. 125 mL wine of 9% alcohol by volume ≌1 unit alcohol). Migraine was deemed present only when two or more of the following specific symptoms supporting this diagnosis as proposed by the International Headache Society  were reported: unilateral headache, associated nausea/vomiting, photophobia, pulsating in nature or visual aura.
The presence of headache at any time in the perioperative period was documented as a dichotomousdependent variable. Univariate analysis determined the association between the explanatory variables and perioperative headache. Using these data, an initial ‘core’ model of perioperative headache was constructed from the variables with P < 0.1. This included age, gender, history of migraine, frequency of previous headache and alcohol consumption. The duration of fasting from food and fluid and caffeine consumption was also included because these are commonly thought to be determinants of perioperative headache. The other variables were successively added to the model and logistic regression performed by the backward stepwise method, eliminating those with P > 0.05. None of the additional variables were retained in the core model. Backward stepwise regression was continued on the core variables until the final model was identified. Analysis was conducted using SPSS® v.9 for Windows® (SPSS, Inc., Chicago, IL, USA).
Fully completed data sheets were obtained from 208 of 230 patients (90%). Operations included the specialities of gynaecology (38%), urology (29%), orthopaedics (19%) and general surgery (14%). All patients received propofol for the induction of anaesthesia and nitrous oxide for maintenance. All patients requiring endotracheal intubation (5% of the total) received atracurium or vecuronium. Patients received various opioids and volatile agents (Table 1). Forty-four (21%) patients had taken analgesics for headache in the month before the operation, but none had done so within 12 h of surgery. The overall incidence rate of perioperative headache was 28% (n = 59), with 24% (n = 49) of the study population experiencing preoperative headache. A smaller number (10%, n = 20) of patients experienced postoperative headache. Four per cent (n = 10) of patients experienced their first headache in the postoperative period. Although 49 patients experienced headache preoperation, in 38 (76%) it had disappeared postoperation. The association between the explanatory variables and perioperative headache is shown in Table 1. The significant factors in the multiple regression model are outlined in Table 2.
Our multivariable model of perioperative headache in day case surgery found that only the frequency of previous headache and low alcohol consumption were clearly identified as risk factors, with a history of migraine being a weak association. Increased frequency of previous headache (one per month or more) as a predictor of perioperative headache was also found in earlier studies of perioperative headache [2,4].
Our finding that low weekly alcohol consumption increases the incidence of perioperative headache is new and surprising. It cannot be explained by a withdrawal phenomenon because perioperative headache was unrelated to the duration since the last alcohol consumption. Whether patients with moderate, regular alcohol consumption have a degree of hepatic enzyme induction, with faster clearance of anaesthetic drugs, is unknown. However, our data and that of others [2,5,6] found no link between the anaesthetic technique or the duration without food or fluid, although a prolonged duration of fluid fasting was implicated in one study .
Our model may suggest a weak non-significant association between a history of migraine and perioperative headache. Migraine sufferers plausibly may be at particular risk of perioperative headache because stress and anxiety or prolonged starvation may be precipitating factors. However, recent studies have overlooked this issue, despite the remarkable changes in surgical and anaesthetic practice over that period. Two-thirds of our migraine sufferers who developed perioperative headache were female. It has been reported that the female gender increases the likelihood of headache per se and that females may be 2.3 times more likely to report postoperative headache . Our study did not find gender to be an independent risk factor when adjusted for other confounding variables. This may reflect cultural differences in symptom reporting or differences in anaesthetic practice.
Caffeine withdrawal has been implicated as a risk factor for perioperative headache. Indeed, some groups have advocated prophylactic administration of intravenous caffeine .
However, a decade ago, premedication, thiopental induction and endotracheal intubation were much more commonly used than today, even in day case procedures. Modern anaesthesia tends to avoid premedication, propofol induction is almost universal, ‘balanced’ analgesia is often employed and the use of the laryngeal mask is commonplace. We found no association with total reported caffeine consumption or duration without fluids that would be linked to caffeine withdrawal duration, suggesting that, contrary to popular belief, this is not an important factor. The choice of anaesthetic technique did not affect the incidence of perioperative headache.
Most day case surgical patients are unlikely to volunteer that they are experiencing headache unless directly questioned in the pre- and postoperative periods. Therefore, these risk factors and the presence of headache should be specifically sought during assessment both pre- and postoperation, with a view to offer simple analgesic therapy. In addition, it may be reassuring to know that a significant proportion of patients with a preoperative headache will not wake up with a headache after surgery. Such measures may increase patient satisfaction with day case anaesthesia and surgery.
In conclusion, in a multivariate model of predictors of perioperative headache in day case surgery, we identified an increased frequency of previous headache and the low alcohol consumption as reliable predictors, and a history of migraine as a weak association. Caffeine withdrawal and duration of fasting from food or fluids were not risk factors for perioperative headache.
1. Hecht AD. Creating greater efficacy in ambulatory surgery. J Clin Anaesth
1995; 7: 581-584.
2. Myles PS, Hunt JO, Moloney JT. Postoperative minor complications. Anaesthesia
1997; 52: 300-306.
3. Headache Classification Committee of the International Headache Society. Classification and diagnostic criteria for headache disorders, cranial neuralgias and facial pain. Cephalgia
1988; 8 (Suppl 7): 19-22.
4. Nikolajsen L, Larsen KM, Kierkegaard O. Effect of previous frequency of headache, duration of fasting and caffeine abstinence on perioperative headache. Br J Anaesth
1994; 72: 295-297.
5. Faithfull NS. Post-operative headache — multifactorial analysis. Eur J Anaesthesiol
1991; 8: 59-63.
6. Goodwin AP, Rowe WL, Ogg TW, Samaan A. The effect of shortening the pre-operative fluid fast on postoperative morbidity. Anaesthesia
1991; 46: 1066-1068.
7. Weber JG, Klindworth JT, Arnold JJ, Danielson DR, Ereth MH. Prophylactic intravenous administration of caffeine and recovery after ambulatory surgical procedures. Mayo Clin Proc
1997; 72: 621-626.