Sir,—Hildick-Smith and colleagues1have reported high complication rates after pacemaker upgrade, with 45% of patients suffering one or more complications. We were initially surprised by this rate and were prompted to review the experience of surgically upgrading pacemakers at our hospital, which implants approximately 500 new pacing systems each year.

Between 1983 and December 1997, 74 patients’ pacemakers were surgically upgraded from a single chamber (either AAI or VVI) to a dual chamber system. Forty five per cent of the upgrades were performed for pacemaker syndrome or worsening cardiac failure, 16% for atrioventricular (AV) block in patients with AAI pacemakers, 7% for carotid sinus hypersensitivity, 5% for miscellaneous reasons, and 27% were coincident with elective generator replacement. Nine per cent of these patients developed a wound or generator pocket infection requiring antibiotic treatment, 17% suffered a lead displacement or failure, and 15% required their upgrade pacemakers to be explanted (predominantly because of persistent infection or generator erosion). Therefore, 36% of patients suffered one or more complications, which is comparable to the 45% reported by Hildick-Smith et al.1

Our patients needing surgical reintervention were younger (58.5 (21.3)v 71.8 (12.9) years, p = 0.009) but otherwise had the same personal and operator characteristics, and pacemaker generator sizes as those without complications, albeit with a tendency to a lower body mass index. Infection was the predominant predictor of requiring further surgery (odds ratio 16.3, 95% confidence intervals 1.8 to 145.1). Complication rates for patients whose pacemakers were upgraded coincidentally with generator replacement were not significantly different from the remainder of the patient group.

These findings support the conclusion of Hildick-Smith et al that pacemaker upgrade should not be done in the absence of a firm indication. Atrial or dual chamber pacing should be the primary procedure wherever possible, as subsequent upgrade has a high morbidity. Recent prospective evidence strongly supports atrial based pacing in patients with sick sinus syndrome,2-4 if not in those with AV block. We await the results of further trials in patients with AV block,5 but it is clear that pacemaker upgrade should be avoided where possible, and certainly should not be performed opportunistically in the asymptomatic or uncomplaining patient. The onus is to select the correct pacing mode in the first instance.