Portal de Periódicos da CAPES

Patient-Controlled Analgesia: Proxy-Controlled Analgesia?

2008; Lippincott Williams & Wilkins; Volume: 107; Issue: 1 Linguagem: Inglês

10.1213/ane.0b013e31817532ae

1526-7598

Elliot J. Krane,

Cardiac Arrest and Resuscitation

There was a time when all analgesia in the hospital was controlled by proxy, that is, not by the patient but by someone acting in behalf of the patient. Indeed, probably from the time of the Crimean War, until very recently, a nurse decided when a patient needed a dose of an opioid analgesic, usually without firm criteria any more specific than "as needed for pain." Those days were not the good old days; both the medical and nursing literature were replete with papers showing that patients frequently did not receive adequate analgesia due to many barriers to its effective delivery. But, nearly 40 yr ago, a new methodology first reported in this journal revolutionized the administration of parenteral analgesia, patient-controlled analgesia (PCA),1 and, in the ensuing decades, PCA has become the convention for the administration of parenteral opioids to adults and older children.2 The safety mechanism inherent in PCA is that, as a patient self-administers small opioid doses, they may become progressively sedated and will sleep before their blood opioid levels trespass far into the range of toxicity, at which point the patient ceases their button-pushing demands for more opioids. There are two facets of the available technology that bypass this inherent safety mechanism: the use of continuous "basal" infusions of opioids and the practice of PCA by proxy, that is, the activation of the demand button by someone other than the patient, on their behalf. The former, basal opioid infusions, are infrequently used in the routine management of adult pain because experience has shown a higher incidence of oxygen desaturation and respiratory morbidity associated with basal infusions.3–5 However, continuous opioid infusions are often used to augment PCA in pediatric pain management practices,6 and the data regarding adverse effects of background infusions in children are less convincing and inconsistent.7–11 The second method of bypassing the safety of PCA is PCA by proxy, and this is also a common practice in pediatric centers.12 Pediatric PCA by proxy, the subject of the report of Voepel-Lewis et al. in this issue of Anesthesia & Analgesia,13 is predicated on the desire to make PCA technology and its effective analgesia available to those patients either too young to effectively activate the dose demand button, or to those too physically or developmentally impaired to do it. PCA by proxy, which may refer to button activation either by a nurse or a family member, is controversial because, by substituting the proxy's judgment for the patient's perceived need and "demand" for opioid, the risk of inappropriate opioid administration and attendant toxicity is created. Furthermore, PCA by proxy has been associated with several fatalities of adult patients (although it should be pointed out that many more adult fatalities have been associated with PCA not-by-proxy), and thus was the subject of a Joint Commission on Accreditation of Healthcare Organizations (JCAHO) Sentinel Event Alert in which the JCAHO recommended that PCA by nurse-proxy be used only after the establishment of patient selection criteria and monitoring standards, and that family members not be allowed to serve as the patient's proxy.14 In the report by Voepel-Lewis et al.,13 a retrospective chart review of the records of 302 children, approximately half received PCA by nurse-proxy superimposed upon a basal opioid infusion. In this series, PCA by proxy was associated with lower pain scores, similar sedation scores and arterial oxygen saturations, and a lower incidence of events requiring dose reduction, oxygen, or stimulation. However, more children in the by-proxy group (7.6%) experienced "rescue events," that is, required either mechanical means, tracheal intubation, or medication to counter airway obstruction, hypoventilation, and/or hypoxemia. Overall, if one combines the patients in each group who required an intervention of any kind (both "threshold" and "rescue" interventions), the conventional PCA and the PCA by proxy groups are indistinguishable. What are we to conclude from this and how may we reconcile these findings with those of Anghelescu et al.,15 who reported in this journal in 2005, the results of an even larger retrospective survey of pediatric PCA by proxy (in which the proxy was either or both nurse or parent), and found no higher incidence of adverse events or rescue interventions? The first conclusion one may derive is that PCA and PCA by proxy are in fact quite safe in the pediatric population. Combining the data of these studies, we have nearly 2000 subjects in whom analgesics were administered by the patients themselves or by a proxy, and are reported to have an incidence of respiratory effects that is much less than commonly reported to occur in adults using PCA.16 Furthermore, one sees confirmation from the Voepel-Lewis et al. study that PCA by proxy is in fact the method of PCA used for younger and sicker patients. Indeed, the by-proxy group had an incidence of concomitant respiratory disease 150% greater, neurologic disease 300% greater, and cognitive impairment 10-fold higher than the comparator group treated with PCA. Naturally, a retrospective chart review cannot easily separate the confounding clinical variables from proxy use of PCA as the cause of the increased incidence of rescue events, and one wishes that Voepel-Lewis et al. had statistically analyzed the effect of these factors in their two treatment groups by performing a multivariate analysis. Therefore, their conclusion that PCA by proxy is associated with increased risk is unconvincing, for one can reasonably speculate that postoperative opioid use per se is associated with respiratory complications in children with cognitive impairment, respiratory disease, and neurologic disease. It seems most likely that in children of similar morbidities (Anghelescu et al.'s patient population did not include children with significant cognitive impairment) PCA by proxy may not be associated with inordinate risk while it, indeed opioid use in general, may pose a clinically significant risk of airway obstruction, oxygen desaturation, and hypoventilation in children with cognitive, neurologic, or pulmonary comorbidities. Having read these studies, and other smaller series that describe PCA by proxy in children, I remain convinced that, in this higher risk group of children, PCA by proxy is no more or less dangerous than conventional old-fashioned nurse-administered opioid analgesia, probably the original form of PCA by proxy since the days of Florence Nightingale, and the most commonly used alternative method of IV analgesia available. The task that lies ahead is convincingly to demonstrate or refute this conclusion by conducting well controlled prospective clinical trials that control for comorbidities and age, standardize the dose regimens, analyze the potentially adverse use of concomitant basal opioid infusions in high risk populations, and to continue to develop and test both new drugs and technologies that facilitate analgesic delivery in novel and safe fashions.17 In the meanwhile, PCA by proxy is here to stay in most pediatric medical centers,12 and, as long as it is, its prescribers would be wise to heed the conditions set forth by the JCAHO14: that it be used with criteria for appropriate patient selection, and that patients be carefully monitored, giving consideration to the routine use of electronic patient monitoring. To these recommendations, I would add that analgesic doses delivered by the proxy be triggered by specific criteria, such as pain scores. Finally, individual practitioners and medical centers will need independently to decide whether to diverge from the JCAHO's recommendation that parents or guardians be routinely prohibited from serving as their child's proxy in the absence of evidence that this is an unsafe practice. To abandon PCA technology altogether in impaired and young children will be a return to the "bad old days" of inadequate analgesia and needless suffering.