Introduction

Cancer pain afflicts millions of people worldwide every year, yet it can be well or completely controlled in 80%–90% of patients (1–3). Exactly how many of the estimated 6.6 million people worldwide who died from cancer last year (4) experienced pain at any one time is difficult to ascertain. The reasons for this are discussed below. Nevertheless, in spite of the major advances in pain control over the last 15 years, cancer-related pain continues to be a major international public health problem (5–21).

Although pain is recognized as an extremely common symptom in patients with cancer, studies to date show a wide variation in the reported prevalence (22–24). Three main factors influence this:

1. Difficulty in making generalizations to different health care settings or different patient groups because of variation in the design of prevalence studies (25,26)

2. Inherent difficulties in assessing the presence or absence of pain, particularly because no one “gold standard” assessment system exists, exacerbated by attempts to grade the severity of cancer pain in a variety of ways

3. Difficulty in defining the type of pain (27), because pain associated with cancer has features of both chronic and acute pain and can be either the direct or indirect result of the cancer (28,29)

Pain and cancer are not synonymous (30). Evaluation of pain in advanced cancer is primarily clinical and is based on pattern recognition. Attention to detail is necessary to prevent inappropriate treatment (31). Comprehensive pain assessment is important, but initial treatment with analgesics should not be withheld until these have been carried out (30).