Between December 1989 and March 1996, more than 6000 patients were tre
ated with patient-controlled analgesia (PCA) at Auckland Hospital. The
overall incidence of potentially life-threatening complications was l
ow (0.28%). A small number (276) received PCA with a background opioid
infusion. This technique was associated with a higher incidence of su
ch complications (1.08%, P < 0.05). To further characterize the safety
and utilization of PCA, a subgroup of 300 patients was analyzed. The
average duration of PCA was 76.4 +/- 39.2 hr, The peak morphine consum
ption was highest on the day of operation (45.4 +/- 37.0 mg) and rapid
ly declined over the next 3 postoperative days (40.6 +/- 39.0, 33.3 +/
- 26.2 and 27.8 +/- 36.6 mg, respectively). The ratio of drug demands
to deliveries decreased from 1.76 on the morning of the first postoper
ative day to 1.17 on the evening of the third. The percentage of patie
nts with inadequate analgesia (pain score greater than or equal to 3/1
0) and an inability to comply with physiotherapy (Bruggemann comfort s
core less than or equal to 2/10) was high on the first postoperative d
ay (42% and 18%, respectively). Men used significantly more morphine t
han women (141.7 +/- 123.6 versus 102.7 +/- 111.2 mg, P < 0.0001) and
general surgical patients used more morphine than urology and orthoped
ic patients (152.6 +/- 136.9 versus 96.0 +/- 84.2 and 83.7 +/- 97.9 mg
, P < 0.001). There was no association between morphine consumption an
d age (r = -0.216). Of the 6% of patients who experienced hypoxemia an
d 2% who experienced respiratory depression, virtually all had one of
three risk factors: bolus dose greater than 1 mg morphine, age greater
than 65 years, or intra-abdominal surgery. The most common side effec
ts were nausea and sedation. The incidence of nausea was highest on da
y 1 (28%) and decreased over the next 2 days (14.3% and 4.7%, respecti
vely). A similar pattern was observed with sedation (incidence over th
e first 3 days: 28%, 9.3%, and 3.3%, respectively). Overall patient sa
tisfaction scores were high (8.3/10 +/- 1.9). We conclude that the ris
k of serious complications with PCA is very low, but worrying degrees
of hypoxemia and bradypnea do occur We suggest prescribing regimens th
at may reduce complications and identify patients at high risk. (C) U.
S. Cancer Pain Relief Committee, 1997.