“…Possible major sources of error, such as bias, procedural changes and different investigators (Loan, Morrison & Dundee, 1968) were overcome using double-blind method, strict adherence to the protocol and the use of the same investigating personnel. Errors arising from the patient population (Loan, Morrison & Dundee, 1968) Time after administration (h) Figure 4 Mean relief scores in patients with initially severe pain: 0, dipyrone; 0, pethidine.…”