US Pharm. 2017;42(3):HS24-HS27.

ABSTRACT: The management of surgical and nonsurgical pain in the hospital setting frequently involves the use of patient-controlled analgesia (PCA) that is delivered by a programmable infusion pump. The concept of PCA centers on giving patients control over their own pain management. Significant interpatient variability and patients’ safety concerns about opioids can make the selection of safe, effective doses challenging for clinicians. A rational framework for the selection of patients, analgesic agents, starting doses, and subsequent dose adjustments is essential for reducing risk and optimizing pain management with PCA.

Patient-controlled analgesia (PCA) refers to the delivery of analgesics immediately upon patient demand. The most commonly recognized modality is IV PCA administered through an infusion pump.1 The concept of PCA, however, can be extended to any method of administration, including subcutaneous, epidural, inhaled, nasal, and oral, provided that drug delivery is administered on demand upon patient request. Infusion-pump technology has advanced considerably over the years, but the general framework is structured around preprogrammed settings for a loading dose, demand or bolus doses, a lockout interval (time that must elapse between demand doses), an hourly or 4-hour dose limit, and continuous or basal infusion.1 Clinicians may prescribe bolus-only PCA, especially for patients who are opioid-naïve, or they may use any combination of these settings.

Although PCA can provide effective pain relief, the selection of appropriate PCA candidates and dose settings requires careful consideration to avoid serious risks such as respiratory depression and opioid overdose. The scope of this article will be limited to the use of IV PCA in the inpatient setting for the management of both surgical and nonsurgical pain.

Evidence for PCA

The use of PCA has gained in popularity since the first commercially available infusion pump was introduced in 1976.2 A large body of research has compared the use of PCA with more conventional analgesia-delivery methods, such as nurse-administered IM bolus doses. Outcomes of interest include differences in analgesic effect, patient satisfaction, total opioid consumption, patient safety, incidence of side effects, postoperative complications, cost savings, and length of hospital stay.

A frequently cited meta-analysis by Ballantyne and colleagues compared as-needed IM dosing and PCA in postoperative adult patients from 15 randomized, controlled trials.3 The PCA group demonstrated greater analgesic efficacy by 5.6 on a scale of 0 to 100, a difference that was small but statistically significant (P = .006). The meta-analysis showed greater patient satisfaction when PCA was used (P = .02), as well as trends toward reduced opioid consumption and shorter length of hospital stay in patients using PCA.3 Data from other studies that evaluated outcomes with PCA have been conflicting or inconclusive.1,4 Although the evidence base for PCA warrants critical evaluation, it is important to balance this with the subjective, individual nature of pain and the complexities of real-world pain management that may not be captured in controlled trials.

Selecting Appropriate Candidates for PCA

PCA is commonly used to manage postoperative pain, and a significant proportion of clinical studies have evaluated PCA in this setting. Patients with acute pain from other causes, such as sickle cell disease and cancer, may also benefit from on-demand analgesia.5,6 Factors to consider when determining whether a patient may safely benefit from PCA include age, cognitive function, physical ability to use the infusion pump, and comorbid conditions.1,4 PCA should not be used in very young children who are unable to understand and follow directions reliably; however, children as young as 4 to 12 years have demonstrated successful PCA use for pain after bone marrow transplantation.7

Reasonable levels of consciousness and cognitive function are required to effectively manage PCA. Although conservative dosing in patients with confusion may prevent a safety event, it is also likely that the pain will not be controlled. Patients should receive PCA education in the preoperative setting rather than immediately after a surgical intervention, when the ability to process directions may be impaired.8 Psychological factors in the absence of confusion and cognitive impairment have the potential to affect successful use of PCA. Autonomy through patient self-determination of the timing of analgesic administration is perhaps the fundamental advantage of PCA. Higher pain scores have previously been correlated with high levels of anxiety. Many patients experience a reduction in anxiety, and therefore have improved analgesia, because of the autonomy of PCA use.9 Conversely, other patients may be fearful to request demand doses—leading to inadequate analgesia—but would accept administration by a trusted professional, such as a nurse or physician.

Comorbid conditions that increase the risk of respiratory depression with PCA include obstructive sleep apnea, morbid obesity, head injury, respiratory failure, renal failure, hypovolemia, and concurrent use of sedative medications, such as benzodiazepines.10 Although these conditions are not absolute contraindications to PCA, conservative dosing and more stringent monitoring are warranted. Sedation always precedes respiratory depression and is a more reliable indicator than respiratory rate. Renal function should be considered in drug-selection and dosing strategies. Fentanyl and hydromorphone are preferred over morphine or meperidine in the setting of renal dysfunction to avoid accumulation of active neurotoxic metabolites (morphine-6-glucuronide and normeperidine).11 Morphine is the most commonly used and studied opioid for PCA, closely followed by fentanyl and hydromorphone. Although methadone and meperidine also may be administered by PCA, the use of these agents is limited by adverse effects and complex pharmacokinetics, and they should be initiated only by a clinician with specialized training.

Dosing Strategies for PCA

A loading dose is an optional clinician bolus given postoperatively or during a pain crisis to bring the pain down to a manageable level. A loading dose may be larger than subsequent on-demand bolus doses such as morphine 2.5 mg, hydromorphone 0.4 mg, or fentanyl 25 mcg.12

A bolus dose should provide clinically significant analgesia, but it should not exceed accepted starting doses if the patient is opioid-naïve. There is no validated method of anticipating opioid requirements in opioid-naïve patients, so close follow-up is essential to determine whether dose titration is necessary. Elderly patients generally require a lower dose of opioid and are at greater risk for respiratory depression compared with younger patients, so they should initially receive conservative starting doses.13 Accepted starting bolus doses include morphine 1 mg, hydromorphone 0.2 mg, and fentanyl 20 mcg.12 One of the main advantages of on-demand bolus doses is that some degree of safety is afforded by a negative feedback mechanism. This means that if a patient were to request sufficient bolus doses to cause sedation—which always precedes respiratory depression—he or she would stop requesting subsequent boluses, thus preventing further clinical deterioration.

The lockout interval prevents repeat bolus administration until a predetermined period of time has elapsed. The patient may press the delivery button multiple times within the lockout period; however, this may be due to confusion, boredom, or anxiety, so this occurrence should not be used as a indication of uncontrolled pain unless an in-depth symptom assessment has been performed. Conversely, some patients report inadequate analgesia but have not used the maximum number of available doses within an hourly period. Although it is appropriate to further educate the patient in such instances, patients may have a maximum natural rate at which they will press the button that will not increase, even in the setting of uncontrolled pain.14 In this situation, a dose increase is indicated. Standard lockout intervals, which range from 6 to 15 minutes, should be tailored based on pharmacokinetic properties of the opioid being used. A shorter interval of 6 to 8 minutes may be use for lipophilic opioids with rapid onset of analgesia, such as fentanyl, whereas a slightly longer interval of 10 to 15 minutes may be used for morphine. The lockout interval should provide adequate time for the opioid to reach full effect, avoiding the risk of dose-stacking and delayed presentation of sedation.15

A basal rate or continuous background infusion should not be initiated in opioid-naïve patients; studies have shown that the addition of a basal rate does not improve the analgesic effect, but increases the risk of side effects and respiratory depression.16,17 If a basal rate is instituted for an opioid-naïve patient based on frequent bolus demands and high opioid requirements, it should be approximately no more than 50% of total daily opioid requirements. A good rule of thumb for determining relative basal infusion and bolus dose settings is that the bolus dose should be 50% to 100% of the hourly rate.12

A basal rate may be added for opioid-tolerant patients when PCA is initiated based on an equianalgesic conversion of chronic, long-acting opioids and a subsequent 25% to 50% dose reduction for cross-tolerance. Patients with cancer pain or pain crisis may receive a higher percentage (80%) of total daily opioid requirements as a basal rate. These patients may also be prescribed higher bolus doses with a longer lockout interval of 20 to 90 minutes compared with opioid-naïve postoperative patients.1,12

A number of approaches should be used when pain is uncontrolled on PCA settings. First, it should be verified that the patient has actually received currently prescribed doses for an appreciable period of time. Drug delivery may be impeded by mechanical issues, obstruction or loss of IV access, or incorrect programming of the pump. Modern PCA infusion pumps have a memory function to record total dose and volume administered, as well as the number of bolus doses both requested and successfully delivered.4 The total amount of opioid delivered over a period of time should be verified using this memory function or nursing flow sheet to guide dose titration.

Caution should be used when a basal rate is initiated and titrated, as this overrides the negative-feedback safety mechanism of PCA. If the patient receives sufficient doses to cause sedation, he or she will continue to receive opioid regardless of bolus requests, greatly increasing the risk of sedation and, ultimately, respiratory depression. The basal rate should be increased by no more than 100% and no more frequently than every 20 to 24 hours. More rapid titration would not allow for assessment of current dose settings at steady state and could lead to delayed drug accumulation, with serious side effects. Bolus doses may be titrated more rapidly—as quickly as every hour—until the patient achieves clinically significant pain reduction. A three-point reduction on a numeric pain scale of 0 to 10 is considered clinically significant.18

PCA infusion pumps can be programmed with hourly or 4-hour dose limits.1 For example, a PCA with IV boluses of morphine 2 mg and a lockout of 10 minutes could be set with an hourly limit of 6 mg. Once the patient has successfully requested three bolus doses, he or she would be unable to receive subsequent boluses until an hour has passed. This can provide a theoretical safety net when bolus or basal doses are titrated; however, it should not be necessary if a comprehensive pain assessment and rational titration are performed. Frequently, the hourly limit is simply calculated as the sum total of opioid made available through the basal rate and boluses doses if the patient were to request a bolus at the end of each lockout interval. Calculating this dose limit can provide a useful double-check for clinicians before initiating dose titrations.

Discontinuation of IV PCA

Clinicians must be mindful of the specific goal of therapy with PCA; patients with acute surgical pain may expect complete resolution of pain and discontinuation of all opioids rather than management of acute or chronic pain crisis and stabilization of a home regimen. As a patient’s pain resolves postoperatively and he or she regains use of the gut, doses may be scaled back and rotated to oral options. If PCA is used to determine opioid requirements for chronic pain or cancer pain, in which the source of pain is unlikely to improve and may in fact worsen, a component of the total daily opioid requirement likely will have to be provided as a long-acting analgesic. Typically, 50% of the total opioid requirement may be provided as long-acting analgesia and 10% to 15% as short-acting analgesia administered at appropriate dosing intervals.12

Whenever a patient is rotated between opioids, it is prudent to perform a 25% to 50% dose reduction to account for incomplete cross-tolerance.12 Patients may have a misconception that IV opioids are more effective based on their faster onset of action and peak effect. Education, reassurance that equianalgesic doses will be provided, and close follow-up should help alleviate these concerns. Oral opioids typically provide more sustained analgesia compared with bolus IV doses of opioids, and this benefit should be communicated to patients.


The autonomy derived from the use of PCA does not guarantee appropriate analgesia or patient satisfaction. Pain management, similar to any other therapeutic area, cannot be successfully approached with a one-size-fits-all mentality. Rather, clinicians must tailor the opioid selection, delivery method, and dose settings to each patient individually, with the expectation that multiple adjustments may be required before the optimal balance is struck between analgesic effect and safety. A conservative approach is always reasonable in the prescribing of high-risk medications such as opioids, and this approach should be applied to PCA. With proper education and hands-on experience, however, clinicians can learn to use PCA effectively for a variety of pain conditions.


1. Grass JA. Patient-controlled analgesia. Anesth Analg. 2005;101(suppl 5):S44-S61.
2. Evans JM, Rosen M, MacCarthy J, Hogg MI. Apparatus for patient-controlled administration of intravenous narcotics during labour. Lancet. 1976;1:17-18.
3. Ballantyne JC, Carr DB, Chalmers TC, et al. Postoperative patient-controlled analgesia: meta-analyses of initial randomized control trials. J Clin Anesth. 1993;5:182-193.
4. Macintyre PE. Safety and efficacy of patient-controlled analgesia. Br J Anaesth. 2001;87:36-46.
5. Citron ML, Johnston-Early A, Boyer M, et al. Patient-controlled analgesia for severe cancer pain. Arch Intern Med. 1986;146:734-736.
6. Davis MP, Weissman DE, Arnold RM. Opioid dose titration for severe cancer pain: a systematic evidence-based review. J Palliat Med. 2004;7:462-468.
7. Dunbar PJ, Buckley P, Gavrin JR, et al. Use of patient-controlled analgesia for pain control for children receiving bone marrow transplant. J Pain Symptom Manage. 1995;10:604-611.
8. Grissinger M. Safety and patient-controlled analgesia: part 2: how to prevent errors. P.T. 2008;33:8-9.
9. Gil KM, Ginsberg B, Muir M, et al. Patient-controlled analgesia in postoperative pain: the relation of psychological factors to pain and analgesic use. Clin J Pain. 1990;6:137-142.
10. Baxter AD. Respiratory depression with patient-controlled analgesia. Can J Anaesth. 1994;41:87-90.
11. Conway BR, Fogarty DG, Nelson WE, Doherty CC. Opiate toxicity in patients with renal failure. BMJ. 2006;332:345-346.
12. McPherson ML. Demystifying Opioid Conversion Calculations: A Guide for Effective Dosing. Bethesda, MD: American Society of Health-System Pharmacists, Inc; 2010.
13. Macintyre PE, Jarvis DA. Age is the best predictor of postoperative morphine requirements. Pain. 1996;64:357-364.
14. Love DR, Owen H, Ilsley AH, et al. A comparison of variable-dose patient-controlled analgesia with fixed-dose patient-controlled analgesia. Anesth Analg. 1996;83:1060-1064.
15. Ginsberg B, Gil KM, Muir M, et al. The influence of lockout intervals and drug selection on patient-controlled analgesia following gynecological surgery. Pain. 1995;62:95-100.
16. Mather LE, Woodhouse A. Pharmacokinetics of opioids in the context of patient controlled analgesia. Pain Rev. 1997;4:20-32.
17. Parker RK, Holtmann B, White PF. Effects of a nighttime opioid infusion with PCA therapy on patient comfort and analgesic requirements after abdominal hysterectomy. Anesthesiology. 1992;76:362-367.
18. Farrar JT, Young JP Jr, LaMoreaux L, et al. Clinical importance of changes in chronic pain intensity measured on an 11-point numerical pain rating scale. Pain. 2001;94:149-158.

To comment on this article, contact