Williams, J.E., Peacock, J., Gubbay, A.N., Kuo, P.Y., Ellard, R., Gupta, R., . . . Ross, J. (2015). Routine screening for pain combined with a pain treatment protocol in head and neck cancer: A randomised controlled trial. British Journal of Anaesthesia, 115, 621–628. 

To compare the effectiveness of a pain management protocol with usual care

Patients who were referred for pain control were screened for worst pain severity on a 10-point numerical scale, and those who scored four or more were randomized to usual care or the experimental protocol group. The protocol treatment was individualized to the patient based on assessments. Protocols included the formulation of a pain treatment plan, weekly reassessments, the provision of an educational brochure, a discussion with a pain control doctor, and a referral to palliative care or other services as needed. Patients were followed for three months. Usual care patients could be referred to the pain control group but were not proactively scheduled for ongoing weekly assessments. Study assessments were done at baseline and at one and two months.

• N = 129
• MEAN AGE = 59 years (range = 19–80 years)
• MALES: 58%, FEMALES: 42%
• KEY DISEASE CHARACTERISTICS: All patients had head and neck cancers, and the majority were cancers of the larynx and pharynx.
• OTHER KEY SAMPLE CHARACTERISTICS: Overall, 87% of participants were receiving weak opioids at baseline, and the average morphine equivalent dose at baseline was 29 mg per day (range = 5–510 mg).

• SITE: Single site  
• SETTING TYPE: Not specified  
• LOCATION: United Kingdom

• PHASE OF CARE: Multiple phases of care
• APPLICATIONS: Palliative care 

Single-blinded, randomized, controlled trial

• Brief Pain Inventory (BPI), the average of four scales as a pain severity index
• Pain Management Index (PMI)
• EuroQol Five Dimensions Questionnaire (EQ-5D)
• Patient satisfaction
• Cost effectiveness analysis

There were no significant differences between the groups in pain measures throughout the study. Patient satisfaction was higher in the study group. The mean cost of pain treatment was significantly higher in the study group. The analysis showed a low probability of the intervention being cost-effective. Pain scores declined significantly over time in both groups.

The protocol did not demonstrate any significant benefit over usual care.

• Baseline sample/group differences of import
• Risk of bias (no blinding)
• Unintended interventions or applicable interventions not described that would influence results 
• Key sample group differences that could influence results 
• Measurement/methods not well described 
• Other limitations/explanation: The sample was underpowered. At baseline, the range of morphine equivalents was much higher in the usual care group. The timing and method of BPI measures was not described. Changes in medications and the use of additional referrals between groups over time was not discussed, so it was unclear if there were substantial differences in the care between groups. The phase of care was not clearly stated. Both groups were managed in the same pain practice, so it was possible that protocol aspects also were applied to usual care patients.

The findings of this study did not show a benefit of the specific protocol approach with weekly reassessments in the clinic as used here. This study did show that pain screening could be implemented and combined with good usual care to reduce pain severity.