Use of opioids for acute pain in hospitalized patients

INTRODUCTION — The use of multimodal, opioid sparing analgesia is a key concept and a guiding principle for acute pain management. Multimodal analgesia involves the use of two or more agents or techniques with differing mechanisms of action. Whereas an overarching goal for pain management is to avoid excessive use of opioids, for patients with or expected to have moderate to severe pain, multimodal analgesia may require the use of opioids.

This topic will discuss the optimal use of opioids for acute pain in opioid naïve patients who are in the hospital. An overall approach to acute pain management, use of opioids for acute pain in the ambulatory setting, and management of acute pain in patients who are opioid dependent are discussed separately.

●(See "Approach to the management of acute perioperative pain in infants and children".)

●(See "Management of acute pain in opioid naïve adults in the ambulatory setting".)

●(See "Management of acute pain in the patient chronically using opioids for non-cancer pain" and "Management of acute pain in patients with opioid use disorder".)

ROLE OF OPIOIDS IN ACUTE PAIN MANAGEMENT — Opioids should be used along with nonopioid analgesic techniques and medications, and only added when those strategies are insufficient. Opioids are associated with short-term side effects (ie, respiratory depression, excessive sedation, nausea and vomiting, pruritus, urinary retention, constipation) and long-term adverse effects (ie, tolerance, dependence, opioid induced hyperalgesia, sexual dysfunction, mood disorders, or withdrawal upon conclusion of therapy) and possible opioid misuse and/or use disorder [1-6]. (See "Risk of long term opioid use and misuse after prescription of opioids for pain".)

Opioids are commonly overprescribed after surgery, and many patients do well with minimal postoperative opioids. Emerging data suggest that nonopioid analgesics may be as or more effective than opioids for treating acute pain in many patients [7-17]. (See "Management of acute pain in opioid naïve adults in the ambulatory setting", section on 'The decision to prescribe opioids'.)

GENERAL PRINCIPLES FOR OPIOID USE FOR ACUTE PAIN — Basic principles for the use of opioids for acute pain are as follows:

●In the setting of multimodal analgesia, opioids should be the last medication to be used and the first to be stopped.

●Functional goals (eg, the ability to rest/sleep, to take deep breaths and cough, and to participate in physical therapy) should be used as endpoints rather than a particular pain score.

●Pain after surgery, injury, or medical conditions can vary widely in the course of the day from minimal pain at rest to severe pain after physical activity. For this reason, we typically prefer to use immediate release, short-acting opioids rather than extended release, long-acting formulations. The more rapid onset and offset of immediate opioids makes them easier to titrate to the need for analgesia, and to taper off opioids as pain subsides. Longer acting opioids may be appropriate for select patients expected to have severe pain for more than a few days (eg, after major surgery, severe trauma). (See 'Choice of opioid' below.)

●Pain after surgery or injury should improve over the first days to weeks. Opioid requirements should likewise diminish.

●Opioids should be used for the shortest period of time possible. Absence of pain should not be the criterion for stopping opioids.

●For patients who undergo elective surgery, preoperative patient education should include expectations for postoperative pain and the appropriate use of opioids. Where appropriate, all patients who receive opioids should understand the plan to stop opioids prior to complete resolution of acute pain, the importance of multimodal analgesia, and that opioids should not be used to treat concomitant non-surgical pain, chronic pain, or as a sleep aid.

OPIOID METABOLISM — Differences in opioid metabolism and excretion may be important considerations when choosing opioids.

●Most commonly used opioids undergo phase I metabolism in the liver via one of the cytochrome P450 (CYP) pathways (CYP3A4 and CYP2D6), with or without phase II glucuronidation. Exceptions include morphine and hydromorphone, which undergo only phase II glucuronidation (table 1). Metabolism of opioids is reduced in patients with significant liver disease, and dosing intervals should be increased to avoid drug accumulation (table 2).

●Opioids metabolized by CYP pathways (eg, oxycodone, fentanyl, hydrocodone, tramadol, oliceridine, methadone) may interact with drugs that either inhibit or accelerate CYP metabolism (table 3 and table 4). In addition, patients may have polymorphisms of CYP genes that affect drug metabolism. Polymorphisms may contribute to either diminished or absent metabolic enzymes or excessive metabolism, either of which can change the clinical effect of a given dose of opioid. Prevalence of CYP polymorphisms varies by ancestry. Variances in metabolism should be considered in patients who don’t respond to pro-drug opioids or have more of an effect than expected. (See "Overview of pharmacogenomics".)

●Some opioids have active or toxic metabolites. For some opioids, metabolites can accumulate in patients with kidney disease (table 5). Recommended dosing for patients with kidney disease appears in the drug interaction program (available through UpToDate) for individual drugs.

Examples of opioid metabolism include:

•Morphine – Morphine is metabolized into morphine-6-glucuronide, which has analgesic effects, and morphine-3-glucuronide, which does not have analgesic properties but can lead to brain excitability and seizures. Both metabolites are excreted in the urine. Thus, morphine should be avoided, doses reduced, or dosing interval extended in patients with kidney dysfunction.

•Codeine and hydrocodone – These are pro-drug opioids; they have only weak analgesic effects, but have metabolites with stronger mu opioid receptor (MOR) activity. Thus, patients who are rapid metabolizers may be at increased risk of respiratory depression or metabolite toxicity. Conversely, patients who have reduced metabolism may experience reduced analgesic benefit from these medications.

•Tramadol – Tramadol is formulated as a racemic mixture that is metabolized via the CYP pathway. While the (+)-enantiomer of tramadol has some intrinsic MOR activity, its primary metabolite (+)-O-desmethyl-tramadol (M1) is a much stronger MOR ligand. Similar to other CYP prodrug pathways, rapid and ultra-rapid metabolizers may experience stronger opioid effects while nonmetabolizers may experience limited efficacy. M1 is renally excreted. Reduce dose or do not use in patients with renal insufficiency. In addition to MOR activity, tramadol also acts as a serotonin reuptake inhibitor ([+]-enantiomer) and a norepinephrine reuptake inhibitor ([-]-enantiomer).

•Meperidine – Meperidine is metabolized to normeperidine (excreted in the urine), which has weak analgesic effects but can precipitate seizures. Meperidine should only be used for short term analgesia, if at all, and is contraindicated in patients with kidney dysfunction. (See "Management of chronic pain in advanced chronic kidney disease" and "Management of chronic pain in advanced chronic kidney disease", section on 'Drugs that should be avoided'.)

ASSESSING PAIN — Visual analog scales or other numeric pain scales are commonly used parameters for opioid orders. While these are helpful to establish objective outcomes and as a measure to assess change after administering a therapy, we find them to be unreliable as the only determinant for opioid use and dosing. We prefer to base opioid use on functional outcomes (eg, resting comfortably, able to breathe deeply and cough, able to complete physical therapy).

Opioids should be held for patients with respiratory depression or who are excessively sedated, regardless of pain level parameters. For example, a patient who reports a pain level at 7 out of 10, but appears sedated or has a low respiratory rate, should not receive further opioid therapy until those side effects have resolved. (See 'Oral dose adjustments' below.)

Acute pain assessment is discussed separately. (See "Approach to the management of acute pain in adults", section on 'Postoperative monitoring for pain control and side effects'.)

INPATIENT OPIOID ANALGESIA

Oral versus intravenous (IV) administration — We suggest using oral rather than intravenous (IV) opioids whenever possible for hospitalized patients unless immediate pain control or rapid titration is necessary. This approach is consistent with guidelines for acute pain management [18,19]. Available data suggest that IV opioids do not provide superior postoperative analgesia compared with oral formulations, even for very painful surgery [20-24]. Oral formulations often cost less than using IV opioids and allow the patient to continue the same drug after discharge if necessary.

IV administration results in rapid rise in effect site concentration leading to faster relief but also may lead to more sedation, increased respiratory depression, and a stronger euphoric effect. In a single institution retrospective database study of over 31,700 patients who were treated with opioids in the emergency department, IV administration was associated with increased adverse events (ie, nausea, vomiting, hypotension, oxygen desaturation) compared with oral administration (10.7 versus 1.8 percent) [25]. Conclusions from this study are limited by lack of data on some comorbidities (eg, obesity, obstructive sleep apnea), lack of data on opioid doses used, and limitations inherent in database studies.

Many opioids can also be delivered via subcutaneous, intramuscular, transdermal, or transmucosal routes of administration. However, utility of these formulations in acute pain settings is limited by variable absorption, delayed onset, pain during administration, and/or cost.

Oral opioids

Scheduled versus as needed dosing — Scheduled dosing may provide more even analgesia but may result in administration of unnecessary opioids and increased risk of excessive sedation or respiratory depression. Thus, for patients with moderate pain who require opioids, opioids should typically be prescribed on an as needed basis. For patients with severe pain, practice differs among contributors to this topic. One contributor prescribes only immediate release (IR) oral opioids, and only on an as needed basis. For patients with severe pain, another contributor may prescribe extended release (ER) oral opioids on a scheduled basis for the first two to three days or longer if necessary, along with as needed IR opioids, and then changes to as needed administration of IR opioids only.

There are little data comparing the use of scheduled versus as needed dosing of opioids for acute pain management in adults, and results of existing studies are mixed.

One single institution parallel group prospective observational study compared outcomes in 105 inpatient medical patients with acute pain who received scheduled versus as needed opioids (oral or IV bolus) [26]. As needed opioid administration was associated with higher mean pain intensity (1.8 ± 2.57 versus 4.1 ± 2.9/10) compared with scheduled administration. Mean opioid consumption was similar in the two groups. Adverse effects consisted of constipation and nausea, and were also similar in the two groups.

Opioid administration for acute pain for patients who chronically use opioids differs, and is discussed separately. (See "Management of acute pain in the patient chronically using opioids for non-cancer pain".)

Choice of opioid — The choice of opioids for acute pain should be determined by patient factors (hepatic failure, renal insufficiency, altered cytochrome P450 [CYP] expression), potential drug interactions (CYP upregulation or downregulation, concomitant selective serotonin or norepinephrine reuptake inhibitors), and patient experience with opioids in the past. Patients should be asked about positive and negative experiences with prior opioid use (eg, nausea, pruritus, poor efficacy) (table 1).

●Duration of action – The expected duration of action for each type of opioid is a rough estimate only. The duration of action is highly variable and is influenced by the dose, variations in metabolism, subjective patient experience, and combination with other therapies. Importantly, use of opioids along with nonopioid analgesics (eg, acetaminophen or nonsteroidal anti-inflammatory drugs [NSAIDs]) is likely to increase the efficacy and duration of action, compared with use of opioids without those medications [27,28]. (See "Nonopioid pharmacotherapy for acute pain in adults", section on 'Options for nonopioid pharmacotherapy'.)

●Immediate release (IR) versus extended release/long acting (ER/LA) formulations – For most patients the author administers IR opioids rather than ER/LA formulations for acute pain in opioid naïve patients [19,29-34]. IR opioids are easier to titrate due to faster onset than ER/LA opioids and are easier to titrate off as pain lessens. However, less frequent dosing and longer duration of action of ER opioids may help patients to sleep.

For some patients with severe pain who require high doses of IR opioids, one contributor to this topic administers a low dose of ER/LA opioid on a scheduled basis to provide a background opioid level for the first few days after surgery, and adds as needed IR opioid to provide an option for self-titration. If ER/LA opioids are used, they should be the first opioids discontinued as pain improves. (See 'Discontinuing opioids' below.)

All patients require close monitoring and dose adjustments, particularly during the first few days after surgery. (See 'Oral dose adjustments' below.)

●Combination formulations – We avoid using combination formulations that include an opioid and acetaminophen or ibuprofen. Separate administration allows administration of the nonopioid medication on a fixed schedule regardless of the patient's opioid utilization, without limitation by the maximum daily dose of the nonopioid. For this reason, we do not routinely use hydrocodone, which is only available as a combination formulation in the United States.

Oral opioid options — Options for oral opioids for acute pain include the following (table 1):

●Oxycodone – Oxycodone is the oral opioid we most commonly prescribe for acute pain. It is typically well-tolerated, inexpensive, and does not require an active metabolite for effect. It is subject to drug interactions because it is metabolized by CYP2D6 and 3A4 to active and inactive metabolites. Oxycodone is excreted in the urine; doses should be reduced and dosing intervals increased in patients with kidney dysfunction. (See 'Opioid metabolism' above.)

Oxycodone is available in 5 mg tablets, which allow convenient dosing as low as 2.5 mg, and is available as an elixir.

Usual dose for opioid naïve patients – 5 to 10 mg orally every 3 to 4 hours as needed.

●Hydromorphone – Oral hydromorphone is also typically well tolerated, does not have active metabolites, is inexpensive, and is available in oral tablet or elixir formulations. Doses should be reduced and dosing intervals increased in patients with kidney dysfunction.

Usual dose for opioid naïve patients – 2 to 4 mg orally every 3 to 4 hours as needed.

●Tramadol and tapentadol – These are mixed mechanism opioids, with both weak mu opioid receptor activity and norepinephrine reuptake inhibition.

•Tramadol also inhibits serotonin reuptake. Tramadol requires CYP2D6 metabolism to have the full mu opioid receptor effect, so there will be patients who report good efficacy and others who report very poor efficacy. It should be avoided in patients taking serotonergic drugs (eg, selective serotonin reuptake inhibitors [SSRIs], serotonin-norepinephrine reuptake inhibitors [SNRIs], or monoamine oxidase inhibitors [MAO-I]) due to risk of serotonin syndrome. Tramadol is sometimes called a "weak opioid", but it is associated with similar adverse effects as other opioids, including risks of misuse and physiologic dependence.

Usual dose for opioid naïve patients – 50 to 100 mg orally every 4 to 6 hours as needed.

•Tapentadol has no active metabolites. It is not available as a generic formulation so tends to be more expensive than other options.

Usual dose for opioid naïve patients – 50 to 100 mg orally every 4 to 6 hours as needed.

●Morphine – Morphine is available in IR and ER formulations as well as an elixir. The lowest available IR tablet dose is 15 mg, which is a large starting dose for an opioid naïve individual. If using morphine IR tablets, we recommend splitting the tablet in half for a 7.5 mg dose. Morphine is associated with histamine release, pruritus, and sedation. It should be avoided in patients with kidney dysfunction due to potential accumulation of active or toxic metabolites. (See 'Opioid metabolism' above.)

Usual dose for opioid naïve patients – 7.5 to 15 mg orally every 3 to 4 hours as needed.

●Codeine – We avoid using codeine because of wide interpatient variability in metabolism, unreliable analgesia, and related incidences of adverse events (see 'Opioid metabolism' above). In a 2015 review of 39 prior reviews (total of 467 studies) that evaluated the analgesic efficacy of single doses of 41 different analgesic drugs administered for postoperative pain, codeine was the least effective analgesic studied, with a number needed to treat for 50 percent pain relief of 12 (95% CI 8.4-18) for codeine 60 mg [15].

Usual dose for opioid naïve patients – 15 to 60 mg orally every 4 hours as needed, no more than 360 mg in 24 hours.

●Fentanyl – Fentanyl is available in transmucosal (buccal and sublingual) formulations. Transmucosal formulations are absorbed rapidly and can be titrated to pain needs, though dosing is imprecise. These formulations may be useful for patients with time limited pain (eg, dressing changes), and as an alternative to IV administration in patients who cannot take oral medication (eg, with head and neck cancer) or with limited absorption (eg, short gut syndrome).

Usual dose for opioid naïve patients:

•Buccal lozenge – 200 mcg over 15 minutes, may repeat once 30 minutes after the start of the lozenge as needed; subsequent dose may be administered ≥4 hours after initial dose

•Buccal tablet – 100 mcg, may repeat once after 30 minutes; subsequent dose may be administered ≥4 hours after the initial dose

●Methadone – Methadone is a long acting mu opioid receptor agonist commonly used for pharmacotherapy for opioid use disorder (OUD), but it can also be used for pain control. When used for analgesia, it is dosed at approximately every eight hours, whereas it is dosed once daily for OUD. Some clinicians administer low-dose methadone for opioid naïve patients who undergo major surgical procedures and are expected to have severe pain and prolonged hospital stay. It is available in both oral and IV forms.

Methadone is unique among opioids in that it is also an antagonist at the N-methyl-D-aspartate (NMDA) receptor, which may improve analgesia compared with other opioids. Intraoperative use of methadone has been shown to improve post-surgical pain and reduce post-surgical opioid use in the first 24 hours after surgery [35,36]. There are little data available for its use in acute pain management in opioid naïve patients.

Disadvantages of methadone include the following:

•It can prolong the QT interval and has been implicated as a cause of drug induced torsades de pointes [37].

•Methadone has variable bioavailability and can be difficult to convert to and from other opioids and from IV to oral methadone. Calculating morphine milligram equivalent doses of methadone is not reliable at higher doses.

•Methadone is subject to drug interactions, as it is metabolized by cytochrome P450 enzymes.

•Methadone has a long elimination half-life with significant interindividual variation, with reported half-life as short as 8 hours and as long as 87 hours after a single dose (mean half-life ranged from 15 to 43 hours with most in the 20 to 30 hour time frame) [38]. The analgesic duration is highly affected by redistribution due to lipid solubility of the drug, and is therefore typically shorter than the elimination half-life of methadone. Analgesic duration has been reported as 4 to 8 hours after a single oral dose and may increase to 8 to 12 hours after repeated dosing [39,40]. With little data on the use of oral methadone for post-surgical pain in opioid naïve individuals, this wide observed variation of elimination half-life makes dosing recommendations for acute pain in opioid naïve individuals challenging.

Due to the complexities associated with methadone use, it should only be prescribed by clinicians familiar with its use and should usually not be continued after discharge from the hospital.

Usual dose for opioid naïve patients – 2.5 to 5 mg orally every 8 hours.

●Buprenorphine – Buprenorphine is a mu opioid receptor agonist with high affinity for the receptor but low intrinsic activity, and is an antagonist at the kappa-opioid receptor. It is often used for pharmacotherapy for OUD, but may also be used to treat both acute and chronic pain.

The doses used for pain are typically much lower than those that are used for OUD. As an example, the usual dose of the buccal film in opioid naïve patients is 75 mcg, versus a starting dose of 2 to 4 mg sublingual for OUD. For acute pain, buccal film, transdermal patch, and IV/intramuscular (IM) formulations are available.

Buprenorphine is associated with less opioid-induced hyperalgesia and may produce less respiratory depression than other long-acting opioids. Clinical experience using buprenorphine for acute pain is limited but increasing, and studies of various preparations used for acute pain after different types of surgery have generally found efficacy similar to other opioids.

•A 2018 meta-analysis of 28 randomized trials that compared intravenous, intramuscular, or sublingual buprenorphine with morphine for acute pain found no statistical differences in pain scores, respiratory depression, sedation, or time to rescue analgesia, despite the prolonged half-life of buprenorphine [41]. Buprenorphine caused less pruritus.

•A 2023 systematic review found 15 randomized trials that compared transdermal buprenorphine with other analgesics (transdermal fentanyl, tramadol, various NSAIDs) or placebo used for acute postoperative pain [42]. Pain control with transdermal buprenorphine was similar to transdermal fentanyl, while results comparing buprenorphine with other analgesics were mixed. The certainty of evidence that buprenorphine improves analgesia was very low, due to high or unclear risk of bias and the few studies available for each comparator.

Usual dose for opioid naïve patients:

•Buccal film – 75 to 150 mcg buccal every 8 to 12 hours

•Transdermal patch – The optimal doses and timing of placement for the transdermal patch have not been determined. Doses used in studies have been from 10 to approximately 50 mcg per hour, often placed from 6 to 48 hours preoperatively, with the patch left in place for up to one week [43].

Use of buprenorphine for acute pain in patients with OUD is discussed in detail separately. (See "Management of acute pain in patients with opioid use disorder", section on 'Patients taking buprenorphine' and "Management of acute pain in patients with opioid use disorder", section on 'Buprenorphine for acute pain'.)

Oral dose adjustments — It is not uncommon for patients to report unsatisfactory pain control or bothersome opioid related side effects when on opioid therapy.

●For lack of sufficient analgesia, assuming nonopioid analgesics have been optimized, it is important to determine whether the analgesic level is inadequate all of the time or if the analgesia wears off prior to the next dose being due. (See 'Assessing pain' above.)

•If the patient does not experience sufficient relief at all after taking the opioid, it is reasonable to increase the dose as a first step (eg, increase oxycodone from 5 mg to 10 mg).

•If relief occurs but does not last long enough, keep the dose the same and decrease the dosing interval.

•If modest dose and/or interval adjustments do not provide adequate pain relief, options include switching to a different oral opioid at an equipotent dose, or administering IV opioids, rather than escalating further the dose of the same drug.

•For patients who continue to have severe pain, consider performing an epidural or peripheral nerve block (if not already in place), or where available, consult an acute pain service for consideration of advanced analgesic strategies (eg, ketamine or lidocaine infusion). (See "Nonopioid pharmacotherapy for acute pain in adults".)

●For opioid-related side effects (eg, nausea, pruritus, constipation), we recommend switching to a different opioid, adjusting dose, or stopping opioid therapy altogether.

IV opioids — IV opioids may be administered by intermittent clinician bolus, patient controlled analgesia (PCA), or in select circumstances by continuous infusion. Continuous IV infusion of opioids is rarely indicated in opioid naïve individuals other than when intubated or in end-of-life care and should only be used in a monitored setting (eg, intensive care unit [ICU]) with continuous pulse oximetry and end-tidal carbon dioxide monitoring. (See "Pain control in the critically ill adult patient", section on 'Opioid analgesics'.)

Patient controlled analgesia (PCA) versus clinician-delivered intermittent bolus — Intravenous opioid boluses can be delivered by a nurse or by the patient with a PCA device. The PCA pump is programmed to allow the patient to self-administer small doses of opioid with a fixed lockout interval (table 6).

PCA is a widely used modality for IV opioid administration in patients who are awake and capable of understanding and controlling the device. PCA reduces delay in access to pain medication compared with as needed clinician boluses and provides the patient more control of the timing of their doses.

However, IV bolus dosing rather than PCA may be useful in some settings, including the following:

●Patients who have breakthrough pain while taking oral opioids.

●The first night after total knee arthroplasty, as it is not uncommon for a patient to get behind and experience severe pain when the spinal anesthetic wears off.

In this setting we typically cancel the IV opioid order the morning after surgery and use only oral opioids thereafter.

●The first night after surgery for patients who have postoperative nausea and vomiting (PONV), and who will take oral opioids once PONV resolves.

In a 2015 meta-analysis of 49 randomized trials (3400 patients) that evaluated PCA without a background opioid infusion versus clinician-administered methods of opioid administration (ie, IV bolus or intermittent infusion, intramuscular, subcutaneous, oral), PCA improved patient satisfaction and reduced pain scores slightly (9 to 10 points less on a scale of 0 to 100) [44]. PCA modestly increased 24-hour opioid consumption (7 morphine milligram equivalents, 95% CI 1-13 mg), and increased the incidence of pruritus (15 versus 8 percent of patients), without a difference in other adverse effects, including respiratory depression.

Patient controlled analgesia

Pump settings — There is wide inter-institution variability in PCA pump settings, without consensus on optimal settings. An example is shown in a table (table 6).

PCA devices are commonly programmed to deliver the following (table 6):

●A demand dose

●A lockout interval until the next demand dose can be given; should be long enough to allow the opioid to reach effect prior to allowing additional dosing and is different for various opioids

Many PCA devices allow for clinician-administered doses to be given through the PCA device in addition to demand doses. With some devices the clinician-administered dose counts towards the maximum total dose programmed into the PCA pump.

The routine use of a continuous background infusion for PCA is not recommended and should be limited to carefully selected patients who are opioid tolerant, receiving end-of-life care, and/or receiving care in an ICU. A continuous infusion delivers opioid whether or not the patient is in pain and whether or not the patient is awake. Continuous infusion is associated with increased risk of respiratory depression in adults, but may be less of a concern in children.

●One retrospective study comparing a fentanyl PCA with or without basal infusion found that there was no difference in reported pain between groups despite significantly larger 24 hour opioid totals in the basal infusion group, but the incidence of adverse events was higher in the basal infusion group [45].

●In a meta-analysis of 14 randomized trials (796 patients, 674 adults, 122 children/adolescents), the addition of a background infusion to demand only PCA increased the risk of respiratory depression in adults (8.2 versus 0 percent, odds ratio 10.18, 95% CI 2.97-34.89) but not in children or adolescents [46]. Conclusions from this study are limited by the relatively small number of events and variable definition of respiratory depression.

●In another meta-analysis of 14 randomized trials of patients <21 years of age (including three of the studies in the previously described meta-analysis), the addition of a continuous infusion did not improve pain scores or opioid consumption, and did not increase adverse events [47]. The quality of evidence was judged to be very low.

Choice of opioid — The most commonly used opioids for PCA are morphine, hydromorphone, and fentanyl (table 6). Choice of opioid should be based on patient factors (eg, kidney or liver dysfunction, side effect tolerance) and hospital protocols, drug availability, and clinician familiarity. We prefer to use hydromorphone for most patients, since it is longer acting than fentanyl, and unlike morphine, can be used in patients with kidney dysfunction and does not cause histamine release.

Small randomized trials have found similar efficacy, side effects, and patient satisfaction with morphine versus hydromorphone PCA after surgery [48,49].

Fentanyl may be used in patients with liver or kidney disfunction or intolerance of morphine and hydromorphone.

Remifentanil, sufentanil, and alfentanil (where available) are not used for postoperative analgesia because of their very short or ultrashort duration of action. Meperidine should not be used for PCA due to potential accumulation of normeperidine and the associated risk of seizures.

Oliceridine is an opioid that was approved by the US Food and Drug Administration in 2020 for short term in-patient IV use, including PCA. Oliceridine is a G-protein selective agonist at the MOR with limited B-arrestin activity which, theoretically, may preserve analgesic efficacy while reducing respiratory and gastrointestinal tract related side effects. Phase III clinical trials demonstrated superior efficacy to placebo and similar efficacy to morphine [50,51]. Head-to-head prospective trials are still needed to compare efficacy and side effect profiles of oliceridine with other commonly used IV opioids.

PCA dose adjustments — PCA doses and intervals can be adjusted to meet the needs of the patient. For patients who are too sedated, or have evidence of respiratory depression, opioids should be held until these side effects resolve. Alternative nonopioid therapies should be considered and, if continued PCA use is required, doses should be decreased and/or lockout intervals increased.

We use the PCA history in addition to pain assessment to titrate PCA orders. For pain assessment we prefer assessment of functional pain parameters (eg, ability to cough, deep breathe, rest comfortably) rather than numerical pain scores (see 'Assessing pain' above). PCA use data can be obtained from review of the PCA device history and/or nursing record. The PCA history should provide the number of attempted patient actuations (records when the button is pushed even if a dose isn’t given due to device lockout), the number of delivered doses, and the total dose of medication delivered.

When adjusting PCA orders for pain it is usually preferable to change one parameter at a time (ie, demand dose or lockout interval) rather than both. In general, when adjusting PCA orders to improve analgesia or function, we consider the following:

●For the patient who gets relief from the demand dose, but repeatedly activates the pump towards the end of the lockout interval, decrease the lockout interval (eg, reduce from 12 minutes to 10).

●For the patient who does not get relief from the demand dose (consistent repeated actuations), increase the demand dose.

Other aspects of the PCA history may be useful as well. As an example, if a PCA history shows no patient PCA attempts overnight but frequent attempts when the patient awakens, possible explanations may include the following:

●If in the morning the patient is showing signs of pain (eg, grimacing with movement and doesn't want to cough), the patient may have adequate analgesia at rest, but increased pain with movement. This could also mean the patient got behind on self-administered dosing when asleep and now has severe pain. A clinician administered dose can be helpful to catch up without needing to change PCA settings.

●If in the morning the patient moves easily, has a strong cough and uses incentive spirometry, or is sedated, this may indicate PCA use for reasons other than pain such as patient anxiety. In this case, we typically reduce the dose and/or increase the interval while maximizing nonopioid therapies.

For opioid-naïve patients who do not receive relief with modest adjustments to PCA dose and interval, additional options include the following:

●Ensure nonopioid analgesic strategies are optimized (eg, local therapy as appropriate [ice, elevation, compression], acetaminophen, NSAIDs).

●Change opioid to different type at equipotent dose.

●Evaluate for surgical or other causes of pain (eg, compartment syndrome, infection, anastomotic leak causing peritonitis, etc).

●Consult the acute pain service for possible regional anesthesia techniques, ketamine or lidocaine infusion. (See "Nonopioid pharmacotherapy for acute pain in adults".)

Clinician boluses — Clinician-administered IV dosing can be useful for patients who are not able to use a PCA device. Since nurses usually administer clinician boluses, orders are typically written for intervals of at least one to two hours (table 7). We do not endorse the use of IV opioid boluses solely as a sleep aid at night.

Transition from IV to oral opioid — Transition to oral opioid formulations should occur as soon as the patient can tolerate oral medications. Nonopioid analgesics (eg, acetaminophen and NSAIDs) should be continued while switching from IV to oral opioids.

When switching from PCA to oral opioids, the following is a reasonable strategy used by one contributor to this topic:

●Discuss the plan to convert to oral opioids with the patient.

●Assess past 24 hour opioid use; convert to equivalent oral opioid dose.

●Prescribe 50 percent of the total equivalent dose, divided into 6 oral doses every 3 to 4 hours as needed.

●Leave the PCA in place for 6 hours after the first oral opioid dose is administered, reassess usage at that time, and titrate oral opioid as necessary.

●Discontinue PCA device.

Discontinuing opioids — For all patients who receive more than a few doses of postoperative opioids, a plan for tapering and stopping opioids should be in place. For most patients who require postoperative opioids or opioids after injury, they can be reduced within two to four days. Nonopioid analgesics should be continued while tapering and stopping opioids.

For patients who receive opioids round the clock for very painful surgery or injury, after a few days we change to as needed administration, and suggest creating a schedule for tapering by progressively reducing the opioid dose and increasing the dosing interval. A reasonable strategy is to reduce the total opioid dose by 20 to 30 percent every 1 to 2 days until cessation. A taper to discontinuation may reduce the risk of opioid withdrawal and may avoid chronic opioid use. Patients who are discharged with an opioid prescription should receive specific instructions for reducing and stopping opioid use.

OPIOID SIDE EFFECTS AND COMPLICATIONS — Opioid related side effects occur in as many as 79 percent of patients who receive opioids for acute pain [52,53].

Common side effects and complications — Common side effects include constipation, sedation, and nausea and vomiting. Other side effects include dizziness, weakness, fatigue, urinary retention, pruritus, hypotension, delirium, and respiratory depression.

The first step to avoid opioid-related adverse events is to limit the use and dose of opioids. Changing the type of opioid may help with some side effects (eg, nausea or pruritus), but is less likely to relieve constipation. We prescribe stool softeners/laxatives and antiemetics for all patients who receive opioids.

Respiratory depression — Opioid induced respiratory depression (OIRD) is possible in any patient who receives opioids, and can result in severe morbidity or death. The incidence of OIRD after surgery is unknown; the reported rate varies widely depending on the definition used. Examples of studies of OIRD are as follows:

●In one literature review, the incidence of respiratory depression in patients using patient controlled analgesia (PCA) after surgery varied from 1.2 percent when assessed by observed respiratory rate, to 11.5 percent using oxygen desaturation by pulse oximetry [54].

●In a prospective international study of 1335 patients who received parenteral opioids and were monitored for a median of 24 hours with continuous capnography and pulse oximetry, 46 percent of patients had at least one respiratory depression event [55]. Respiratory depression was defined as respiratory rate ≤5 breaths/minute, oxygen saturation ≤85%, or end-tidal carbon dioxide ≤15 or ≥60 mm Hg for ≥3 minutes; apnea lasting >30 seconds; or any respiratory opioid-related adverse event.

●Using another indicator of respiratory depression, in one study naloxone was administered to up to 0.7 percent of patients who received postoperative morphine [56].

●In a meta-analysis of 29 studies (1914 patients) that compared PCA to other methods of postoperative opioid administration (eg, intramuscular, IV bolus, subcutaneous), there was not a significant difference in the incidence of postoperative respiratory depression (2.3 versus 2.0 percent, respectively) [44].

●An analysis of closed claims associated with respiratory depression due to postoperative opioid administration found 92 claims, of which 77 percent resulted in severe brain damage or death [57]. Eighty eight percent of these events occurred within 24 hours of surgery and 97 percent were judged to have been preventable.

Risk factors for OIRD include opioid naïve status, older age, obstructive sleep apnea, severe obesity, postoperative status, receipt of other sedating medications, male sex, congestive heart failure, and smoking [55,58-63]. Enhanced monitoring (eg, continuous pulse oximetry, continuous capnography, and telemetry) should be considered for patients with these risk factors. Concomitant use of other sedating medications (eg, benzodiazepines) should be avoided.

Monitoring — Despite the known risks associated with OIRD after surgery, the optimal methods for monitoring patients receiving systemic opioids have not been established. The Anesthesia Patient Safety Foundation (APSF), several other medical societies, and regulatory agencies have issued guidance with varying degrees of specificity for patients receiving parenteral opioids [18,34,58,64-67]. Though oral opioids are increasingly used for inpatients who have acute pain, there are no standards for the monitoring of patients receiving oral opioids.

We follow the APSF recommendation that all patients receiving parenteral opioids be monitored by continuous pulse oximetry and telemetry, in addition to interval nursing assessments of vital signs, pain, sedation, and respiratory function. In addition, we use continuous pulse oximetry and telemetry for at least 24 hours for patients with severe obesity or severe sleep apnea (obstructive and/or mixed) who are receiving oral opioid therapy.

Examples of recommendations for monitoring patients who receive opioids include the following:

●The Center for Medicare and Medicaid Services (CMS) guidelines for the monitoring of patients receiving opioids specify that hospitals must have policies and procedures related to the use of high-alert medications, including IV opioids for post-operative patients [64]. These policies must address patient risk assessment, monitoring frequency and duration, and monitoring methods. At a minimum, vital signs, pain level, respiratory status, and sedation level must be monitored.

●The APSF recommends that all hospitalized patients receiving parenteral opioids should be monitored by continuous pulse oximetry, preferably with centralized telemetry, in combination with intermittent nursing assessment [65,66]. Patients who require supplemental oxygen should also be monitored with capnography or other modalities that measure the adequacy of ventilation.

●An independent international multidisciplinary consensus statement on prevention of opioid related harm recommended that all hospitalized patients receiving opioids should be assessed for sedation using a validated sedation scale at appropriate and repeated intervals [34]. The recommendations did not define appropriate intervals.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Acute pain management".)

SUMMARY AND RECOMMENDATIONS

●General principles

•Opioids should be used as part of multimodal analgesia, should be discontinued prior to nonopioid analgesics, and should be used for the shortest period of time at the lowest effective dose.

•Patients should be educated on the expectations for postoperative pain and appropriate use of opioids, including functional goals, adequate analgesia rather than absence of pain, and the plan for stopping opioids. (See 'General principles for opioid use for acute pain' above.)

●Treating pain in the postanesthesia care unit (PACU)

•Oral opioids can be used for patients who have moderate pain.

•For patients with severe pain, intravenous (IV) opioids are typically used to achieve rapid analgesia.

●Opioids in the hospital – We suggest using oral rather than IV opioids whenever possible in patients who do not require immediate pain control or rapid titration (Grade 2C). IV opioids may cause more side effects than oral formulations and do not provide superior postoperative analgesia. Oral formulations cost less and the same drug can be continued after discharge if necessary. (See 'Oral versus intravenous (IV) administration' above.)

●Oral opioids

•Choice of opioid – For most patients we suggest using immediate release (IR) opioids rather than extended release/long acting (ER/LA) opioid formulations (Grade 2C). IR opioids are easier to titrate to the level of pain due to faster onset, and are easier to taper off as pain lessens. However, ER/LA opioids are a reasonable option for patients with severe pain, and their less frequent dosing and longer duration of action may help patients sleep. If ER opioids are used they should be the first opioids discontinued as pain improves (table 1).

Methadone is a long acting opioid that may be useful for patients with severe pain expected to last several days. (See 'Choice of opioid' above.)

●IV opioids – For IV opioids, we suggest bolus administration (by clinician bolus or patient controlled analgesia), rather than use of continuous infusion (Grade 2C). Continuous infusion is associated with increased risk of respiratory depression, and should only be used for selected patients in a monitored setting (eg, intensive care unit) (table 1).

•PCA versus clinician bolus – Patient controlled analgesia (PCA) reduces delay in access to pain medication, gives the patient control over timing of the dose, and does not increase side effects, including respiratory depression. Clinician boluses may be useful for patients who have breakthrough pain while taking oral opioids or who have a temporary need for IV opioids before using oral formulations. (See 'Patient controlled analgesia (PCA) versus clinician-delivered intermittent bolus' above.)

•PCA opioid – The choice of PCA opioid should be based on patient factors and institutional preference. Efficacy and side effects are similar with morphine versus hydromorphone (table 6). (See 'Patient controlled analgesia' above.)

●Transition from IV to oral opioid – Transition to oral opioid formulations should occur as soon as the patient can tolerate oral medications. Nonopioid analgesics (eg, acetaminophen and nonsteroidal anti-inflammatory drugs) should be continued while switching from IV to oral opioids. (See 'Transition from IV to oral opioid' above.)

●Discontinuing opioids – Opioids should be stopped as soon as possible. For most patients, opioids can be reduced within two to four days. For all patients who receive opioids, a plan for tapering and stopping opioids should be in place. The taper schedule should usually include reductions in both dose and dose interval. (See 'Discontinuing opioids' above.)

●Opioid induced respiratory depression (OIRD) – OIRD is possible in any patient who receives opioids and can result in severe morbidity or death. (See 'Respiratory depression' above.)

•All patients who receive opioids should be monitored for sedation and respiration, and patients who receive IV opioids should be monitored with pulse oximetry with telemetry. Patients with risk factors for respiratory depression (eg, severe obesity, severe obstructive sleep apnea) may require enhanced monitoring.

•Concomitant use of other sedating medications (eg, benzodiazepines) should be avoided in patients who receive opioids.