Purpose:

According to the proposal of the International Headache Society acute postcraniotomy headache (PCH) is defined as a headache of variable intensity, being most serious at the site of surgical intervention and developing within 7 days after craniotomy. Typically, pain resolves within 3 months after surgery. According to the literature the incidence of acute PCH during the postoperative period is 60% among patients undergoing elective craniotomies. No standardized international guidelines are available on the treatment of PCH to the present day. Treatment methods can be different based on the country and institution where they are used, or even physicians individually can have distinct medication regime, which sometimes happens to be habitual and lacks clinical evidences.

The Department of Anesthesiology and Intensive Care Medicine at the University of Debrecen has been giving 100 mg diclofenac with analgesic purpose, as part of the premedication to neurosurgical patients for several years. It was observed that the postoperative headache following craniotomy was milder comparing to data published in the international literature. A pilot study was carried out with similar settings which found the that incidence of PCH was lower and less analgesics were required without any further unwanted side-effects among patients who received 100 mg diclofenac as part of the premedication.

Hypothesis:

100 mg preoperatively given diclofenac significantly reduces the incidence and severity of postcraniotomy headache.

Before surgery, every patient's previous history was taken routinely and along with this, special attention was paid to a history of preoperative headache, its severity and any analgesics used regularly prior to surgery. In this study 200 cooperative adults (over 18 years) undergoing elective craniotomy were randomly assigned into two groups with: 100 patients received 7.5 mg midazolam and 100 mg diclofenac as premedication, while the other 100 members of the control group received only 7.5 mg midazolam one hour prior to surgery. Envelop randomization was used to schedule patients to diclofenac or to placebo groups. Neither the anesthetists, nor the physician obtaining the postoperative VAS scores were aware of the actual patient's grouping (diclofenac or placebo) during the entire study. For induction of anesthesia, we used propofol (1-2,5 mg/kgBW), whereas for maintenance of anesthesia we used the combination of fentanyl, rocuronium, and sevoflurane. Craniotomy and opening the dura occurred always after the tissue saturation of sevoflurane and low-flow anesthetic technique was applied at a sevoflurane concentration ranging from 1.2 to 2.1 V% as required for maintenance of adequate anesthesia. Fentanyl was administered as an initial bolus of 100-150 µg followed by a maintenance dose of 2 µg/kgBW/h. The total dose of fentanyl was calculated offline after the procedure in all cases. The surgical site was infiltrated by a combination of 2% lidocaine and epinephrine in all cases. All patients were admitted to the neurosurgical ICU for postoperative observation. A stepwise analgesic regimen was decided prior the study. Severity of pain was assessed regularly by the ICU nurses and additional analgesics were used if the patient complained a headache severity of VAS >3. In these cases additional paracetamol (1-2 g given intravenously) was administered. If this was not effective, an additional dose of 100 mg diclofenac was given to the patients. In case of further ineffectivity, tramadol was administered in a dose of 100 mg per os or iv. At the end of the study the cumulative dose of analgesics that was administered to the patients was calculated. For sake of clarity we transformed analgesic doses to equianalgetic doses of intramuscular morphine according to the table of the Oxford Pain Site and these values were used for further comparisons of analgesic requirements.

Pain assessment occurred at bedrest state by using a 0-10 scaled visual analogue scale (VAS). Assessment of actual VAS scores was regularly performed by the nurses of the neurosurgical ICU during days 1-5 and rescue analgesics were given according to the actual needs of the patients. An independent physician (who was unaware of randomization) performed an additional check of VAS scores by asking the patient, what was her/his worse pain during the day. As the patient's satisfaction was considered to be dependent on the worse pain, only these VAS scores were used for further analysis.

The investigators intended to answer the following study questions:

• Our research intended to examine whether 100 mg preoperatively give diclofenac reduces the incidence and severity of PCH on the first and fifth postoperative days.
• Is there a difference in the incidence and severity of PCH between patients who underwent infra- or supratentorial interventions?
• Does 100 mg preoperatively give diclofenac reduce intraoperative and postoperative analgesic needs?
• Does 100 mg preoperatively give diclofenac increase the incidence of postoperative hemorrhagic events?

Grouping of the patients:

The anesthesiologist on duty assigned the participants randomly to the control or to the diclofenac group by using envelope randomization technique the day before the surgery.

Members of the control group received 7.5 mg midazolam, while members of the diclofenac group received 7.5 mg midazolam and 100 mg diclofenac for premedication.

Gathering data:

Since no objective method exist to measure pain patients were asked to rate their headache subjectively with the help of a visual analog scale, where 0 stands for no pain and 10 for intolerable pain. In order to make the interpretation of data easier the scale was divided into three parts: between 1 and 3 the headache is mild, from 4 to 6 it is moderate, while 7 or a greater value represents a particularly strong pain.

The sites of the surgical intervention, intra- and postoperative analgesic needs, postoperative complications were recorded individually onto a personal data sheet.

Each data mentioned above was registered by a third pre-trained, independent (blind) member of the study team, who was unaware of the premedication and grouping of patients.

Statistical methods:

Statistical analysis was performed using Statistica for Windows (Statsoft, Tulsa, USA) statistical program. A normality test revealed that data of VAS scores and analgesic requirements show non-normal distribution, therefore a non-parametric Mann-Whitney test was used for further comparisons. A p-value of <0.05 was accepted as statistically significant during analysis.

Inclusion Criteria:

• Patients who were over the age of 18, able to cooperate and underwent elective craniotomy at the Department of Neurosurgery of the University of Debrecen.
• Patients with full awareness and without existing sensory or severe motor aphasia at admission or postoperatively.
• After written informed consent and written agreement was obtained from each participant.

Exclusion Criteria:

• Patients under the age of 18, patients who were unable to cooperate and who underwent traumatic or emergency interventions.
• Patients taking non-steroidal anti-inflammatory agents prior to surgery for any reasons were excluded from the study.