A small study of the treatment of chronic painful diabetic neuropathy (22) exemplifies this approach. In this nonrandomized trial, pain was classified as superficial, muscular, or deep in 75 patients with diabetes mellitus who had chronic, painful diabetic neuropathy and had experienced symptoms for longer than 12 months. Each level of pain was treated differently: Imipramine, with or without mexiletine, was given for deep pain; stretching exercises and metaxalone, sometimes supplemented with piroxicam, were prescribed for muscular pain; and capsaicin was given for superficial pain. Patients receiving this stepped approach were compared with a control group receiving usual care. Among patients receiving the stepped treatment approach, 21% became free of pain, 66% showed improvement, and 13% were unchanged at the end of 3 months. On a graphic pain rating scale of 0 (no pain) to 19 (worst pain imaginable), patients reported scores that decreased by a mean (±SD) of 5.2 ± 0.7, 6.8 ± 0.7, and 6.4 ± 0.6 points in the superficial pain, deep pain, and muscular pain groups, respectively; in contrast, scores did not change appreciably in the control group. All three intervention groups had significantly less pain at the end of the 3 months compared with the control group (P < 0.001).