Frasco, Peter E. MD; Sprung, Juraj MD, PhD; Trentman, Terrence L. MD
A number of reports in the late 1990s suggested that a large percentage of patients continued to experience moderate to severe pain after surgery (1) and that, after surgery, an individual has a 50% to 80% chance of having unrelieved pain (2,3). In 1995, both the American Society of Anesthesiologists (4) and the American Pain Society (5) suggested that quality assurance and improvement programs must be implemented to institutionalize pain management. Several years later, with input from the American Pain Society, the Joint Commission for Accreditation of Healthcare Organizations (JCAHO) created a new pain initiative to ensure appropriate assessment and management of pain in hospitalized medical and surgical patients (2,6–8) New pain management standards were published in August, 1999, and all JCAHO accredited health care organizations were required to comply during the year 2001 (8,9). According to these standards, health care organizations have a responsibility to develop processes to help support improvements in pain management, including methods to ensure: a) recognition of patients’ right to appropriate pain assessment and management; b) appropriate assessment of the severity of pain; c) regular pain assessment, recording, and follow-up; and d) establishment of policies and procedures that support the appropriate prescription of pain medications (10). In addition, the JCAHO suggested that health organizations adopt pain management strategies that align with the principles of quality improvement. Although JCAHO did not explicitly state how to achieve this aim to have some measurable estimate of pain level, pain management programs must include one of the existing pain measurement tools, such as a visual analog scale (VAS) or a verbal numeric pain intensity scale (NPIS) rating. In addition, the education of physicians, nurses, allied health professionals and patients regarding the importance of treating pain in the hospitalized patient was to be a priority for health care organizations.
Although JCAHO has a significant influence on health care organizations, little is known about how compliance with any of the JCAHO’s directives, initiatives, or guidelines influences patient outcomes or costs. The increased organizational emphasis on perioperative pain management, derived from the new JCAHO guidelines, might lead to increased opiate usage, which in turn could lead to increase in adverse outcomes (respiratory depression, respiratory arrest, nausea, emesis etc). Furthermore, if these opioid-induced adverse effects prolonged postanesthesia care unit (PACU) length of stay, organizational expenses and patient charges could have increased.
This study was conducted to assess the effects of the JCAHO pain initiative in our institution. We studied the use of opiates in the perioperative period, PACU length of stay, and complications that may have been related to pain management before and after JCAHO pain initiative. We hypothesized that adherence to the JCAHO pain management standards would lead to increased immediate postoperative opiate use, increased incidence of postoperative nausea and vomiting (PONV) requiring treatment, increased naloxone use for treatment of opioid-induced respiratory depression, and prolonged PACU stay.
In the spring of 2001, a task force including anesthesiologists, internists, pharmacists, and nursing executives implemented the JCAHO pain initiative standards throughout the Mayo Clinic Hospital (MCH) Scottsdale. In compliance with the JCAHO pain initiative, the mandatory use of an 11-point (0–10) NPIS for assessment of pain in PACU was implemented in MCH Scottsdale in September 2001. In addition, response to analgesic therapy was recorded using the same pain intensity scale. Low pain scores were added to the standard discharge criteria. On discharge from the PACU, a pain score was recorded and reported to the next level of care. This step was added to ensure continuity of pain management. Of note, before the implementation of these changes in response to the JCAHO pain initiative, there was no formal policy for the management in pain in hospitalized patients in our facility.
The significance of pain control was enhanced by educating perioperative care nurses and physicians about the concept of pain as the “fifth vital sign” (6). The JCAHO resource book for 2002 was distributed to all nursing units including the PACU. Using this publication and a series of newsletters from JCAHO, a simplified policy outline was constructed. A 3-h pain class was required for all registered nurses, including the PACU staff. The content of the class included pain medication clinical pharmacology, pain assessment using the NPIS, patient-controlled analgesia (PCA), and epidural analgesia techniques. In addition, the nurses were made aware of the literature (11) regarding the relationship between VAS and the severity of pain. Ninety-two percent (519 of 564) of hospital-based nurses completed the training. Before 2002, the only education that nursing staff completed was a 20-min pain station at a mandatory education fair. Finally, the administrative nursing staff performed walking rounds in all units, including the PACU, to ensure that the new pain management practices were followed. After these administrative rounds, a selective audit of documentation of pain intensity demonstrated compliance rates between 90% and 94%.
We reviewed the impact of the new JCAHO pain initiative on the amounts of opioids used in the PACU, opioid-induced side effects (incidence of PONV requiring treatment, use of naloxone for respiratory depression), as well as PACU discharge times. With IRB approval, we accessed records of 541 surgical patients before the JCAHO pain initiative (January 1, 2000 through December 31, 2000) and 541 after the pain initiative (January 1, 2002 through December 31, 2002). This population represents patients from each of the top nine surgical diagnostic related groups (DRG) at MCH. The charts were randomly selected by computer from a complete list of patients and procedures performed within these DRGs for the years 2000 and 2002. The nine DRGs were 148/149 (major small and large bowel procedures), 209 (major joint/limb reattachment procedures of the lower extremity), 334/335 (major male pelvic procedures), 358/359 (uterine/adnexal procedures for non-malignancy), and 499/500 (back/neck procedures not including spinal fusion). Only patients having general anesthesia were included in the study. There were no age restrictions imposed. Analysis of the cases in the two groups revealed no significant differences in the types of cases within the specific DRGs. The following information was extracted from the charts: demographics (age, gender, type of surgery), intraoperative opiate use, PACU opiate use, PACU length of stay, PACU discharge NPIS scores for year 2002 (in 2000 these scores were not recorded), intraoperative and postoperative ketorolac use, perioperative use of antiemetics, and postoperative use of naloxone. Dosages of fentanyl (0.1 × μg dose), sufentanil (0.75 × μg dose), meperidine (0.15 × mg dose), hydromorphone (7 × mg dose), and oxymorphone (10 × mg dose) were expressed as morphine equivalents (ME) in milligrams using these standard potency conversions (12,13). The PACU length of stay was recorded from the time of admission to the time when discharge criteria were met. Readiness for discharge from the PACU was determined by the Mayo Modified PACU Scoring System (MMPDS). The MMPDS is comprised of five criteria (activity, respiration, systolic blood pressure, consciousness, and oxyhemoglobin saturation). Each of these criteria is scored as 0, 1, or 2. An MMPDS score more than or equal to 8 (maximum of 10) was required for discharge from the PACU. The requirements for discharge were unchanged between study periods except for the addition of the requirement for pain scores to be low, to be recorded on the discharge summary, and to be reported to the next level of care. The attending anesthesiologist was notified if the pain was inappropriately intense for discharge. At that time, the patient was evaluated. This evaluation included interview, assessment of vital signs, level of sedation, and pupillary size. Additional medications were given as indicated by this assessment.
Use of antiemetic therapy (droperidol, ondansetron, dexamethasone, promethazine, and hydroxyzine) both intraoperatively and in the PACU was recorded. Antiemetic treatments were divided as follows: 1) antiemetic prophylaxis (intraoperative administration), 2) antiemetic rescue (all patients who received prophylaxis but required a second dose of any antiemetic drug for PONV), 3) antiemetic treatment (patients who had PONV and required treatment in the PACU), 4) postoperative antiemetic (all patients who received antiemetic in postoperative period either as a rescue or as treatment), and 5) perioperative antiemetic (all patients who received treatment at any time in the immediate perioperative period). Charts were excluded from review if 1) peripheral nerve blocks or central neuraxial blocks were used in addition to general anesthesia, 2) if any time data were missing, 3) if the patient was admitted directly from the operating room to the intensive care unit or intermediate care, and 4) if pain score could not be obtained from the patient (e.g., because of unconscious patients or language difficulties).
The primary outcome measures were postoperative dose of opiates (measured in ME) and PACU recovery time between the two study periods. Secondary outcomes were use of PCA in the PACU, frequency and nature of use of antiemetics, and use of naloxone to treat respiratory depression. Populations were compared with Wilcoxon’s ranked sum test, χ2 tests, and the Hodges-Lehmann method. P values were considered significant if P < 0.05. All statistical analyses were performed by using the StatXact 5 software.
Patient demographics are summarized in Table 1. There were no significant differences between patient age and gender in years 2000 versus 2002. Also, patient ages in all five DRG distribution categories were comparable (all P > 0.05). Overall use of ME increased (P = 0.001), primarily because of a large increase in opiates consumed in the PACU (63%) (P < 0.001, Table 2). The intraoperative doses of ME between the two periods did not differ, except for slightly more use of ME in the general surgical DRGs in 2002 (P = 0.01). However, postoperative opiate use increased in 2002 for each DRG category except back surgery. The amount of increase ranged from 34% for orthopedic surgery, to 97% for general surgery. The change in pain management practices between the two study periods was also reflected in increased PCA use beyond the PACU in 2002: 108 patients in 2000 and 138 patients in 2002 (P ≤ 0.05) received PCA before discharge from the PACU. Interestingly, frequency of ketorolac treatments increased significantly from 34.9% in 2000 to 44.9% in 2002, almost exclusively because of an increased in intraoperative use (Table 2). The mean postoperative pain score on discharge from the PACU in 2002 for all groups was 2.8 ± 1.9 (median, 3; mode, 3; interquartile range, 1–4). We cannot report pain scores from 2000, as they were not a part of PACU discharge criteria at that time.
Between the two study periods, overall PACU recovery times decreased by <10% because of decreased PACU stay for urologic (Table 3, 13%, P < 0.01) and back operations (18%, P < 0.001).
Overall incidence of PONV requiring treatment in the PACU was 31.2% in 2000 and 27.5% in 2002; however, in 2002 significantly more patients received intraoperative anti-PONV prophylaxis (Table 4, 53.2% versus 73.8% in 2000 and 2002, respectively). More patients in 2002 received ondansetron prophylaxis compared with 2000 (Table 5, 66.2% versus 34.4%, respectively). In addition, this increased use of ondansetron was significantly different within each of the DRGs in 2002 compared with 2000. In 2000, more patients received droperidol prophylaxis compared with patients in 2002 (Table 5, 9.4% versus 0%, respectively). There was no difference in the use of multidrug PONV prophylaxis among the groups (Table 5). The number of rescue doses (second dose after prophylaxis) of antiemetic was not different between the two study periods, whereas the number of first antiemetic treatments in PACU (primary treatments) was larger in 2000. However, when comparing cumulative rescue treatments and primary anti-PONV treatments in PACU the difference was not significant (169 versus 149 in 2000 and 2002, respectively (Table 4, P = 0.2). The differences in the medications used in the PACU to treat PONV comparing 2000 and 2002 are presented in Table 6. There was no difference in the use of multidrug treatment of PONV in the PACU between the years (Table 6, P = 0.2)
The increased perioperative use of opiates during 2002 did not result in an increased incidence of respiratory depression needing intervention: during the years 2000 and 2002, nine and 11 patients required naloxone in PACU, respectively (P ≤ 1). On discharge from the PACU, an additional 13 patients in 2000 and seven patients in 2002 required naloxone on the surgical nursing unit (P = 0.2).
In October 2002, MCH Scottsdale participated in a survey conducted by JCAHO; a component of the survey assessed compliance with the JCAHO pain initiative. The JCAHO pain initiative was designed to generate adequate patient comfort level or uniformly lower pain scores at rest and with movement in hospitalized patients. Although the JCAHO pain initiative specifies neither the pain measurement tool nor an “acceptable pain score,” our perioperative team set a target of “adequate pain control” before discharge from the PACU. In establishing this policy, the PACU personnel received didactic training and were made aware of literature regarding pain management in the perioperative period. These studies reported that a pain score of ≤4 (on an 11-point scale) is a reasonable goal of analgesic therapy (11,14). The present study was conducted to assess the effects of the JCAHO pain initiative on opioid use in PACU and to assess the potential adverse effects of this treatment (prolonged PACU stay, increase in PONV, increased incidence of respiratory depression requiring naloxone treatment). In our study we demonstrated increased use of opioids in PACU after adoption of new pain standards in 2002 and confirmed low discharge pain scores (median score, 3). Contrary to our hypothesis, the increased opioid use was not associated with either an increase in PACU stay or increased incidence of potential opioid-induced complications. A recent editorial (15) suggested that achievement of pain scores of ≤3 (defined as “minor pain”) by using opioids as the sole predominant method of analgesia could result in an increased frequency of side effects. Vila et al. (16) recently reported an increase in the incidence of opiate-related oversedation or respiratory depression as a consequence of implementation of new JCAHO pain management guidelines. In contrast to Vila et al. (16), we did not note any difference in frequency of respiratory depression requiring naloxone between the two study periods.
Furthermore, increased use of opiates in PACU could be associated with more frequent opiate-induced PONV but this was not confirmed in our study; PONV was present in 31.2% in 2000 and 27.5% in 2002. This incidence of PONV is consistent with reports of others (17–20). At the same time, in 2002 significantly larger number of patients received intraoperative antiemetic prophylaxis compared with 2000, and this increased intraoperative use of antiemetics in 2002 may be a plausible explanation for the unchanged incidence of PONV with increased opioid usage in PACU. We did note a difference in the drugs used for antiemetic prophylaxis and treatment. Patients in 2002 were more likely to receive ondansetron as prophylaxis, whereas patients in 2000 were more likely to have received droperidol. Because our facility discontinued the routine use of droperidol after the “black box” warning from the Food and Drug Administration, droperidol was used in only one patient in 2002, whereas 115 patients received droperidol in either the operating room or in the PACU in 2000. There was no significant difference in the incidence of receiving ondansetron in the PACU; however, more patients received ondansetron as a first-time antiemetic in 2000 compared with 2002.
Before the JCAHO initiative (7) there were several systematic efforts to establish a more comprehensive approach to perioperative pain management (21,22). Bardiau et al. (14) examined the effect of standardization of postoperative pain treatment and reported a significant reduction in postoperative pain scores after introduction of an acute pain service. Interestingly, the use of paracetamol increased whereas the mean morphine administration decreased (14). One of the main determinants in these results was education of the nursing staff in the use of the VAS in the assessment of baseline pain and as a feedback marker for assessing the outcome of analgesic therapy (14).
The JCAHO pain initiative appeared to have improved perioperative pain management without visible adverse effects in 2002. We do not have firm evidence of increased patient satisfaction after implementation of the JCAHO pain initiative, but the pain scores in 2002 that were ≤3 are consistent with “low pain at rest and with movement” (15). Increased use of opioids for comparable types of surgeries between the two time periods suggests that, after the JCAHO pain management initiative, pain management in the PACU was intensified without any clinically significant adverse effects. We did not evaluate the costs associated with compliance with the JCAHO pain initiative nor could we establish a cause-effect relationship between the JCAHO initiative and some change in our practice. For example, increased use of antiemetics in 2002 cannot be tied primarily to the JCAHO pain initiative, as these antiemetics were given intraoperatively; therefore they may reflect changes in anesthesia practice in recent years. Also, the increased use of ketorolac appears to be an institutional trend rather than practice governed by JCAHO regulations. Finally, we identified a larger increase in PCA usage after discharge from PACU in 2002, which may also reflect a practice change—the saving of human (nursing) resources as well as more careful attention to postoperative pain (as a part of the JCAHO pain initiative).
The retrospective nature of the investigation was unavoidable given the nature of the hypothesis; i.e., we were interested in the outcome of a change in policy within our institution. Prospective study was not possible once the policy was implemented. Not all DRGs are represented. The DRGs that were queried represent approximately 15% of the total case volume at our institution. One could argue that use of DRGs rather than actual surgical procedures could lead to differences of the types of surgical cases within the study groups. However, we analyzed the cases performed under the DRGs in the 2000 group versus the 2002 group, and this analysis did not reveal any significant differences in the types of cases within the specific DRGs. For example, within DRG 499/500, the number of single level laminectomies, 2-level laminectomies, and micro discectomies were not statistically different in 2000 versus 2002.
Actual PACU discharge time was equated to the “readiness for discharge.” Although PACU length of stay may have exceeded patient readiness for discharge, in this retrospective study we had no option but to use the recorded discharge time and assume that this was the time when patients achieved stability of all vital signs and adequate pain control. Finally, differences in opioid use may stem from preferential use of narcotic with differential pharmacokinetics during one study period, i.e., using preferentially short-acting narcotics in PACU may theoretically translate into more narcotic usage based on their need for repeat administration. However, at our facility, morphine, fentanyl, and hydromorphone are also used before the end of surgery and during PACU stay. This type of narcotic use was consistent over the years at MCH Scottsdale, including the two periods covered by this study. Also, although conversion factors for calculating narcotic equivalents may differ slightly according to different references (12,13,23), we used the same conversion rate across the two populations in this study. Lastly, we did not monitor respiratory depressions and other events that may have happened beyond 6 hours after discharge from PACU but still may have been related to PACU management.
In conclusion, we detected an overall increase in opioid use in PACU after implementation of the JCAHO pain initiative. This increase was not associated with prolonged PACU length of stay. We also noted increased use of prophylactic antiemetic medications, ketorolac intraoperatively, and PCA postoperatively in 2002, which presumably all reflect changes in anesthesia and perioperative management practices rather than direct JCAHO perioperative-care-driven initiatives.
1. Dolin SJ, Cashman JN, Bland JM. Effectiveness of acute postoperative pain management: I. Evidence from published data. Br J Anaesth 2002;89:409–23.
2. Curtiss CP. JCAHO: meeting the standards for pain management. Orthop Nurs 2001;20:27–30, 41.
3. Warfield C, Kahn C. Acute pain management: programs in US hospitals and experiences and attitudes among US adults Anesthesiology 1995;83:1090–4.
4. Ready LB, Ashburn M, Caplan RA, et al. Practice guidelines for acute pain management in the perioperative setting: a report by the American Society of Anesthesiologists task force on pain management, acute pain section. Anesthesiology 1995;82:1071–8.
5. Campbell J. Presidential Address to the American Pain Society. Los Angeles, CA, 1995.
6. Lanser P, Gesell S. Pain management: the fifth vital sign. Healthcare Benchmarks 2001;8:68–70.
7. Joint Commission releases new pain management standards. Rep Med Guidel Outcomes Res 2000;11:7–10.
8. Berry PH, Dahl JL. The new JCAHO pain standards: implications for pain management nurses. Pain Manag Nurs 2000;1:3–12.
9. Berry PH. Getting ready for JCAHO: just meeting the standards or really improving pain management. Clin J Oncol Nurs 2001;5:110–2.
10. JCAHO begins surveying for new pain standards. Hosp Peer Rev 2000;25:67–8.
11. Collins SL, Moore RA, McQuay HJ. The visual analog pain intensity scale: what is moderate pain in millimeters? Pain 1997;72:95–7.
12. Redmond M, Florence B, Glass PS. Effective analgesic modalities for ambulatory patients. Anesthesiol Clin N Am 2003;21:329–46.
13. Ballantine J. The Massachusetts General Hospital handbook of pain management. Philadelphia: Lippincott Williams & Wilkins, 2002:562.
14. Bardiau FM, Taviaux NF, Albert A, et al. An intervention study to enhance postoperative pain management. Anesth Analg 2003;96:179–85.
15. Berde CB, Brennan TJ, Raja SN. Opioids: more to learn, improvements to be made. Anesthesiology 2003;98:1309–12.
16. Vila H, Nagl PA, Ross TW, et al. JCAHO pain management standards: are patients being harmed [abstract]? Anesthesiology 2003;99:A35.
17. Apfel C, Laara E, Koivuranta M, et al. A simplified risk score for predicting postoperative nausea and vomiting: conclusions from cross-validation between two centers. Anesthesiology 1999;91:693–700.
18. Cohen M, Duncan P, DeBoer D, Tweed W. The postoperative interview: assessing risk factors for nausea and vomiting. Anesth Analg 1994;78:7–16.
19. Gan TJ. Postoperative nausea and vomiting: can it be eliminated? JAMA 2002;287:1233–6.
20. Apfelbaum JL, Chen C, Mehta SS, Gan TJ. Postoperative pain experience: results for a national survey suggest postoperative pain continues to be undermanaged. Anesth Analg 2003;97:534–40.
21. Report on the working party on pain after surgery. Royal College of Surgeons. London: The Royal College of Surgeons of England, 1990;1–36.
22. Quality improvement guidelines for the treatment of acute pain and cancer pain. American Pain Society Quality of Care Committee. JAMA 1995;274:1874–80.
23. Portenoy RK. Contemporary diagnosis and management of pain in oncological and AIDS patients, 3rd ed. Newton, PA: Handbooks in Heathcare, 2000.