Administration of PRN Range Opioid Analgesic Orders for Acute Pain

Abstract

The use of "as needed" or "pro re nata" (PRN) range opioid analgesic orders is a common clinical practice in the management of acute pain, designed to provide flexibility in dosing to meet an individual's unique needs. Range orders enable necessary adjustments in doses based on individual response to treatment. However, PRN range opioid orders have recently come under scrutiny as a source of confusion and as a medication management safety issue. How nurses administer range orders may vary based on their interpretation of the intent of an order, inadequate knowledge of analgesic titration, or exaggerated concerns about opioid safety. The purpose of this study was to investigate nurses' opinions of the appropriate implementation of range orders. Six hundred two nurses from one large academic medical center and one multihospital system completed an online survey using theoretic clinical vignettes to examine their opinions of appropriate analgesic administration practices. The majority of participants chose appropriate responses to the vignettes; however, there was a great deal of variability in responses. Those who had attended pain management courses were more likely to have a higher percentage of appropriate responses than those who had not attended courses. Years in practice and educational level were not significantly related to percentage of appropriate responses; however, there was a trend for nurses with a master's degree to have a higher percentage than nurses with other educational preparation. Consideration of opioid pharmacokinetics can provide logic to develop a new paradigm where range orders are replaced with orders that provide more explicit instructions to titrate an opioid to the most effective dose.

Introduction

Opioids are commonly used to manage moderate to severe acute pain. The amount of opioid required to manage pain is known to vary widely owing to inter- and intraindividual responses to pain and the available opioid analgesics. The use of "as needed" or "pro re nata" (PRN) range opioid analgesic orders is an age-old conventional practice in the management of acute pain, designed to provide flexibility and safety in dosing to meet an individual's unique needs. However, pain is often undertreated because physicians underprescribe opioid analgesics (order inappropriately low doses or prolonged dosing intervals) and nurses administer inadequate doses (give the lowest doses in a provided dose range despite indications that a higher dose is necessary) (Pasero, Portenoy, & McCaffery, 1999).

In 2001, the Joint Commission (JC) incorporated pain management standards into their survey and accreditation process as a means to stimulate improvements in the quality of pain care in the U.S. (Dahl & Gordon, 2002). Simultaneously, a series of national consensus meetings led by the Institute of Medicine fueled the development of a set of national patient safety goals (Chassin, Galvin, & National Roundtable on Healthcare Quality, 1998). Among these goals is a heightened focus on improved communication among caregivers and medication safety, including the use of opioids. The JC Medication Management standard (MM 3.20) states that medication orders are to be written clearly and transcribed accurately and that hospitals are responsible for taking steps to reduce the potential for error or misinterpretation. An important element of compliance with this standard is "there is a documented diagnosis, condition, or indication-for-use for each medication ordered," emphasizing the need for clear understanding and communication between staff (Joint Commission, 2007).

Following the release of the pain management standards and development of national patient safety goals, JC surveyors began inquiring into institutional policy and procedures surrounding the use of PRN range opioid orders. During accreditation reviews, surveyors encouraged detailing in range orders, such as the designation of specific dose or route of analgesia based on pain intensity ratings. The slogan of pain as the "fifth vital sign," originally developed to increase visibility of pain assessment in the clinical arena (APS, 1995), became misconstrued by some as an edict to treat to a target pain rating. With the encouragement of JC surveyors, many institutions developed policies that led to the development of PRN orders based solely on patient report of pain intensity indexed with a numeric scale. An example of this type of order is: "one tablet oxycodone 5 mg orally (PO) for pain less than or equal to 4/10 (scale 0-10); two tablets oxycodone 5 mg PO for pain 5 to 7/10; morphine 2 mg intravenous (IV) for pain 7-8/10; morphine 4 mg IV for pain greater than 8/10." Regrettably, this approach was soon linked to increased rates of significant adverse events (Taylor et al 2003, Vila et al 2005). In one setting (Vila et al., 2005), the incidence of opioid oversedation episodes per 100,000 inpatient hospital days increased from 11 before implementation of a numeric pain rating treatment algorithm to 24.5, a more than twofold increase (p < .001). Although practice guidelines recommend that the choice of analgesic be based on the severity of pain, the route and amount of medication must always be individualized (APS, 2003).

In response to questions and concerns by clinicians about the course of range order policies, the American Society for Pain Management Nursing, in collaboration with the American Pain Society (APS), developed a national consensus statement to support and clarify the use of PRN opioid analgesic range orders (Gordon, et al., 2004). To date, the APS nursing special interest group E-mail listserve, whose members have a special interest or specialize primarily in pain management nursing, (http://mailman.listserve.com/listmanager/listinfo/apsnursingsig) continues to reverberate with ongoing debate and questions about range order competencies and policies. There are at least two sources of common confusion for nurses about a PRN range order. The first is which dose to administer from within a range. The second is the total time interval during which the maximum dose may be administered. In other words, what dose is safe, effective, and within the prescribed parameters? Considerations for dosing have previously been described (Gordon et al., 2004) and include patient and drug characteristics, such as the patient's prior drug exposure, prior responses to analgesics, age, organ function, pain severity, anticipated pain duration, comorbid conditions, and concomitant drug use. The "total time interval" dilemma has also been described (Pasero, Manworren, & McCaffery, 2007). Some nurses may believe the time interval is defined by the time of the first dose, whereas other nurses may interpret the time period as starting from the most recent dose. The latter is referred to as the "rolling-clock" time interval (Pasero, Manworren, & McCaffery, 2007), whereby the time interval ordered serves as the limiting and pivotal guidepost; the maximum cumulative dose should not be exceeded in any block of the ordered time interval.

Management of acute pain, including the administration and titration of range opioid analgesic orders and patient monitoring, is a critical nursing responsibility. Identification of beliefs that result in variable administration of PRN range orders may further facilitate national recommendations about how to construct range orders that will improve patient safety and outcomes in pain management. The purpose of the present study was to document nurses' opinions of the appropriate implementation of a range order. The research questions were:

1. How much agreement is there among nurses about the administration (timing and dosing) of analgesics ordered with a PRN range order?
   
   
2. Do respondent characteristics (such as education, years of nursing experience, pain course attendance) affect knowledge and/or comfort in titration and vignette responses?
   
   
3. What factors do nurses consider when selecting a dose to administer from a range order?

Methods

The study was conducted at one 471-bed academic medical center in the Midwestern U.S. and a multihospital system with five operating units (2,102 total beds) located in a mid-Atlantic state. The study protocol was reviewed and given exemption status by the Institutional Review Boards at both sites. Data were collected from participants using a self-administered internet-based questionnaire that took approximately 10-15 minutes to complete. Risks were considered to be minimal, because data were aggregated via a security-protected public online survey service (http://www.vovici.com/). The internet protocol (IP) addresses were analyzed only in aggregate for the two participating sites; no connection was made between individuals and a computer's IP address or compliance. Participation in the study was voluntary. Respondent confidentiality was ensured throughout all stages of the study. No identifying information including age or gender was collected. All 611 licensed nurses at one site were individually contacted via e-mail with an invitation to participate in the study and were provided a link to enter the web-based survey. Because the second site did not provide individual nurses with email accounts, all nursing staff, totaling 3,323, were invited to participate via a notice posted on the health system home web page.

The questionnaire was developed by the authors and pretested among five staff nurses who completed the questions and then discussed the clarity of the specific items with the authors. The final 15-item questionnaire included items on the following: level of nursing education and practice experience; usual patient population and practice setting; special training, course, or certification in pain management; knowledge/comfort level of opioid titration; patient vignettes for determining timing and dose titration of IV and PO opioids; and the three most important factors to consider when selecting a dose to administer from within a range order. The vignettes are described in the Results section of this article.

Data Analysis

Descriptive statistics were run for respondent characteristics and answers to the vignettes. Each answer to the vignettes was then coded as appropriate (preferred) (given 1 point) or not appropriate (not preferred) (given 0 points). The majority of vignettes were constructed to elicit a response of the soonest next dose based on peak effect of the route administered or the amount of a next dose based on a percentage upward titration. Answers were deemed to be appropriate or preferred if the choice was a dose administered at the time or just after peak effect of the previous opioid dose or if the answer was a dose that had previously been effective or represented a 50%-100% dose titration of a previously ineffective dose. Because of the difficulty in constructing a brief vignette with one correct answer, all but one question had two to three answers that were scored as appropriate. A total score was calculated for each respondent by totaling the number of appropriate answers. In final analysis, the authors could not agree on the preferred response(s) for the dose question in the second vignette and decided to delete that item from the total score analysis (see Titration discussion). Therefore, total scores could range from 0 to 6. Comparisons of responses were used to determine if there were associations between respondent characteristics and responses to clinical vignettes using chi-squared and t tests.

Results

Respondent Characteristics

A total of 602 participants completed the on-line survey in July and August 2007. Response rate was 222 (36%) for the site with individual E-mails and 378 (11%) for the site with only a web page invitation; two respondents did not identify their site. Table 1 summarizes the respondents' characteristics. The largest number of participants was baccalaureate-prepared RNs (56%). The largest percentage (27%) had more than 25 years of experience in nursing. Most (90%) worked with adult patients, with 22% identifying themselves as working with pediatric patients. Medical, surgical, and intensive care units were the most frequent responses for current practice settings. Forty-three percent had previous special training or had attended a course in pain management; however, only nine (2%) held certification in pain management by the American Nurses Credentialing Center.

Although both sites had a policy or written protocol for PRN opioid range orders, when asked if they were aware of one at their facility, slightly fewer than one-half said "yes." Forty-one percent were unsure of the existence of a policy. Nurses were also asked to rate on a 0-10 scale (0 = extremely uncomfortable; 10 = extremely comfortable) their knowledge/comfort level on how to titrate opioids. Titrate was defined as "to determine the effective dose by giving graduated increases or decreases in the amount of a drug." The mean rating was 5.98 (standard deviation [SD] 2.66); however, there was a large amount of variability in responses.

Timing and Titration

Four vignettes were included in the survey ( Table 2 ) for the administration of IV and PO opioid analgesics regarding timing and/or titration of opioids. The first described a situation in which a patient had a fourfold PRN range order for IV morphine (2 to 8 mg every 2 h). Following a first dose of 2 mg at 14:00, respondents were asked to choose how soon another dose could be given. The appropriate answers were 14:15 or 14:30. Sixty-eight percent of the respondents chose one of these answers ( Table 2 ). Twenty-three percent, however, chose to wait the entire 2-hour time period before giving another dose. Respondents were then asked what dose should be given assuming there was no change in pain rating, pain was still severe, and there were no side effects. Seventy-two percent chose 3 mg or 4 mg, which were the preferred answers ( Table 2 ).

The patient described in the second vignette had the same range order (2 to 8 mg IV morphine every 2 h PRN) and had received 2 mg at 12:00 and 3 mg at 13:30 with no relief from either dose. This was followed by a 6 mg dose at 14:00. Sixty-five percent of the respondents chose to wait until 15:30 or 16:00 (which was the appropriate answers) before giving a next dose ( Table 2 ). Twenty-nine percent would have given a dose as early as 14:15 or 14:30. Although a rolling-clock time interval approach would indicate that there is still another 2 mg available in the range order maximum, a 2 mg dose at 14:15 or 14:30 was not preferred, because it was considered to be a previously ineffective dose. At the time of the next dose the patient was described as having reported satisfactory relief with the previous dose of 6 mg, but the pain has returned. The patient is drowsy and arousable, but drifts off to sleep during conversation. Respondents were asked what dose of morphine they would next administer. Nearly one-half (49%) of the respondents chose 4, 5, or 6 mg ( Table 2 ), however, 25% chose none (see Titration discussion). Of note, this item was not included in the total score, owing to author disagreement regarding interpretation of the meaning of the phrase "drowsy and arousable, but drifts off to sleep during conversation."

The third vignette asked participants what they would do if they thought the next dose given to a patient should be less than what was ordered. The order stated "morphine 4 to 8 mg IV every 2 h PRN." The patient had received several doses of 4 mg 2 h apart with "good" pain relief and "no" side effects. Just over one-half (56%) of the respondents said they would call an MD to get an order change, which was the preferred response; however, 22% said they would increase the time between doses (which was considered to be inappropriate), and 22% would give some amount less than what was ordered (which was considered to be inappropriate) ( Table 2 ).

The final vignette presented an order for an oral opioid (oxycodone 5 to 20 mg every 4 h PRN). The patient received 5 mg at 08:00, 10 mg at 09:00, and 5 mg at 11:00. Nurses' opinions were again elicited for the soonest time the next dose could be given and the amount of the next dose if the patient has no change in pain rating and is still in severe pain with no side effects. Fifty-two percent of the respondents chose to give the next dose at 12:00 (preferred) ( Table 2 ). Many (23%) chose to wait a full 4 hours after the last dose, which was given at 11:00, before giving the next dose. Appropriate doses (15 to 20 mg) were chosen by 54% of respondents ( Table 2 ).

If respondents selected an appropriate answer for all vignettes they were given a total score of 6. The average total score for all respondents was 3.64 (SD 1.11). (As mentioned, the dose question in the second vignette was not scored.)

Relationship of Respondent Characteristics and Comfort in Titration and Vignette Responses

Increased comfort level with titration was associated with increased years of experience in nursing (Figure 1).

Figure 1.

Years of experience and comfort in titrating opioids (0 =extremely uncomfortable to 10 = extremely comfortable). *Nurses with <5>25 years of experience (p < .05).

Those with <5>25. Special training or course attendance was also associated with higher titration comfort levels. Respondents that had special training or attended a pain course had significantly higher confidence levels, (6.71 [SD 2.47]) than those who did not (5.43 [SD 2.68]); t = 5.88; p < .001. Comfort level with titration was not associated with total score for preferred answers.

Interestingly, there were no significant differences in total score of appropriate responses by years of experience or education level (Figure 2).

Figure 2.

Years of experience and total score for preferred answers (6 = maximum score, meaning all vignettes were answered with an appropriate or preferred response).

All groups had similar scores. Although not significant, there was a trend for respondents with a master's degree to score higher than other groups. Associate and baccalaureate degree graduates were very similar, and those with licensed/vocational and doctoral preparation tended to score slightly lower. There were no differences in scores between respondents who had previous special training or attended a course in pain management and those who did not. Only nine people were certified in pain management, so comparisons were not meaningful.

Factors to Consider in Opioid Administration

The four factors that were most often chosen as the most important to consider when selecting an opioid dose to administer from a range order were, in order of frequency: sedation level, pain intensity rating, respiratory rate, and the patient's prior response to dosing (Figure 3).

(Enlarge Image)

Figure 3. Respondents were asked to mark the three most important factors to consider when selecting a dose to administer from a range order. Discussion Timing A variable pattern of PRN range order administration is often observed in clinical practice. A range order such as "oxycodone 5 to 20 mg as needed every 4 h" might be administered over 24 hours using incremental doses and varying dosing intervals within any defined 4-hour time period. Although acute pain may vary widely between rest and activity, one must question if a patient's analgesic needs and responses to doses vary in a manner that would require two- to fourfold variations in dosages within a short time period of several days, or if variable patterns of administration more accurately reflect nurses wide range of comfort and opinions of order limits and knowledge of opioid titration. The results of the present study would seem to confirm the latter. It has been suggested (Pasero, Manworren, & McCaffery, 2007) that range orders could be designed similar to IV patient-controlled analgesia (PCA) to afford patients access to rapid administration when needed and minimize delays that can lead to worsened pain. Time to peak effect varies between opioids, ranging from 1 to 5 minutes for IV fentanyl, to 15 to 30 minutes for morphine and hydromorphone. With IV PCA therapy, the time interval or lockout period is set at 5 to 10 minutes for all opioids, allowing rapid maximum, yet safe, usage, because 80% of peak effect characteristically occurs within the first 5 minutes of bolus administration (Upton & Macintyre, 1997). The recommended hourly limit with IV PCA is three to five times the projected hourly requirement (APS, 2003). Thus it would seem rational to construct a PRN range IV opioid analgesic order in a manner that allows a nurse to repeat at least the same bolus dose at least every 15 minutes. Likewise, it would be reasonable to allow PRN doses of short-acting oral opioids as frequently as every hour. Onset of action of a short-acting oral opioid typically occurs within 45 minutes, with peak drug effect in 1 to 2 hours. It is therefore theoretically safe to provide a second dose at peak effect of a first dose of a short-acting opioid, provided the patient has unrelieved moderate to severe pain and there are minimal side effects. Although the use of a "rolling-clock" time interval approach clarifies the timing of PRN orders and facilitates a more consistent interpretation of cumulative dosing, it may hamper needed titration. For example, given an order of oxycodone 5 to 20 mg every 4 h PRN, a patient who has received a dose of 10 mg that resulted in inadequate pain relief would need to wait a full 4 hours before receiving an appropriate increased dose of 15 mg. Although the rolling-clock method provides nurses with a more literal interpretation than alternative time-interval approaches, it is less focused on the pharmacokinetics and pharmacodynamics of the analgesic being administered. Titration Titrate means to determine the effective dose by giving graduated increases or decreases in the amount of the drug. The APS suggests that repeat IV boluses can be used to titrate to a plasma concentration that provides effective pain relief, e.g., morphine 0.03 mg/kg (e.g., approximately 2 mg in a 150-pound individual), or the equivalent dose of a similar opioid every 10 minutes until there is a 50% reduction in pain or the patient reports satisfactory relief (APS, 2003). The total loading dose required can then be considered to be an effective dose for analgesia and made available as a PRN IV bolus dose for subsequent doses (Harris, Kumar, & Rajagopal, 2003). Alternatively, it has been suggested that subsequent doses be increased if at the time of peak effect the patient has minimal pain relief with minimal or no side effects (Mercadante, 2007). Based on pharmacologic observations (Hanks, DeConno, Cherny, Hanna, Kalso, & McQuay, 2001), the next dose can be increased by 50% to 100% for moderate to severe pain and by 25% for more minor adjustments (Pasero et al., 1999). Similarly, oral opioid doses could be titrated as soon as every hour and the total dose required for pain control could then be made available as the PRN dose, or several graduated increased doses could be used to find the most effective dose. This would require a more lengthy order set or use of a pre-established protocol, as described elsewhere (Pasero et al., 2007). For example a one-time order reads: "Give 10 mg PO oxycodone now; if after 1 h pain is unrelieved and there are no side effects give 20 mg PO oxycodone; if 1 hour later pain continues to be unrelieved and there are no side effects give a third dose of 30 mg PO oxycodone." However, multiple orders would be required for slower titration for conditions of unrelieved pain in the presence of side effects and to also stipulate that the dose that was found to be effective is then the only available PRN dose. Beyond understanding the pharmacokinetics of opioid analgesics, titration and use of range orders requires nurses to properly evaluate patient status before administration of an opioid dose. The second vignette in the present survey demonstrated the complexity of decision making when a nurse implements an opioid range order and the need to understand how to lower a dose or titrate down. Recall that the patient in the vignette was given a 6 mg dose and reported "satisfactory" pain relief with a reduction of severe pain to moderate pain. The decision to administer an equal or larger subsequent dose should be an easy one in patients who experience a return of severe pain, provided there are no or minimal side effects. However, the patient in this vignette was described as being excessively sedated and "drifting off to sleep during conversation" at the time of the subsequent dose (which, as noted, resulted in significant author debate and exclusion from the total score). A reasonable and safe action in such a situation is to hold opioid analgesia and any other sedating agents (chosen by 25% of respondents), promptly inform the physician of the patient's status, request an order for nonsedating analgesia, and investigate the possible cause of the excessive sedation. Reducing the dose by 25% to 50% and waiting to administer it until the patient is less sedated are also appropriate actions. Thirty percent chose to administer a dose in this dose reduction range (3, 4, or 5 mg), whereas 24% chose to repeat the previous dose of 6 mg which, although effective for pain, had resulted in a significant side effect. Nurse monitoring of sedation level is frequently cited as an essential function when caring for patients receiving opioids (Pasero and McCaffery 2002, Pasero et al 1999). Indeed, sedation level was cited most often by just over one-half (56.6%) of the respondents when they were asked to choose the three most common factors to consider when selecting an opioid dose. However, this means that nearly one-half of the respondents did not appreciate the importance of sedation assessment and underscores the need for reinforcement of this aspect of safe opioid administration. Relevance of Findings Although one-half to two-thirds of the nurses chose what we considered to be reasonably appropriate responses, the other one-third to one-half chose answers indicating that they err on the side of conservative administration of PRN opioid orders even in the presence of severe pain and minimal or no side effects. This may be explained by being less knowledgeable or comfortable with how to titrate opioids, or by variability in how nurses interpret the time interval. Nearly one-fourth of all of the nurses consistently chose responses that would delay or underdose a patient. Alarmingly, 21% of nurses indicated they would give a dose less than what was ordered without first calling a physician. These findings are of concern on at least two levels. First, this indicates that some nurses may not always follow the basic nursing principle that requires them to contact the prescriber when, for whatever reason, the nurse is unable to follow an order. Second, the decision to give less than allowed by the range order could result in loss of stabilized pain control; recall that the patient described in the third vignette had good pain relief and no side effects from 4 mg of IV morphine every 2 hours. The present findings are similar to those of another study that showed 38% of the nurse respondents thought patients should experience pain before giving a subsequent dose and should not be maintained in a "pain-free" state (Howell, Butler, Vincent, Watt-Watson, & Stearns, 2000). We could find nothing in the literature that described nurses' independent decisions to give less opioid than prescribed in a range order; however, research has shown that nurses will often administer less opioid than allowable when the opioid is prescribed as a PRN dose (Pasero et al 2007, Titler et al 2003). Surveys have consistently shown that nurses tend to select lower PRN doses or withhold analgesics even when a patient's condition warrants the need for more analgesia (McCaffery, Pasero, & Ferrell, 2007). In a study of pain assessment and analgesic administration in elders, Titler et al. (2003) found that the patients in their sample received 25% of the minimum parenteral morphine equivalent of opioid prescribed and that eight patients received no analgesic whatsoever despite having PRN orders for analgesics and demonstrating behaviors indicative of moderate to severe pain. Boer et al. (1997) studied the administration of analgesics that were prescribed in scheduled doses to 150 postoperative patients and found that naproxen and acetaminophen plus codeine were often administered as prescribed. However, patients in the study almost never received the total prescription of parenteral morphine. Nurses cited fears of addiction and respiratory depression and a preference for nonopioid analgesics as reasons for failing to administer the scheduled morphine doses. These findings indicate that nurses should be reminded that the flexibility in dosing made possible by a range order does not include adjusting the dose limits of the order. They should be encouraged to always communicate with other members of the health care team when they think it is necessary to administer more or less opioid than is prescribed. Limitations There are a number of limitations to this study. The survey involved a convenience and voluntary sample of nurses who may not be representative of all nurses. Recruitment was different in the two sites based on E-mail availability. Those who have internet access and/or use E-mail may represent a different group from nurses who do not. A major limitation is the difficulty of using simple vignettes to mirror the complex reality of a patient condition and the comprehensive assessment and decision making nurses perform. No psychometrics analysis was performed in the construction of the nurse survey. Finally, although the authors determined appropriate answers based on onset, peak, and duration of IV and PO short-acting opioids, there were no absolute right or wrong answers. Summary Although the use of PRN range opioid analgesic orders has long been a familiar strategy in the inpatient setting, the safety of this practice has recently been called into question owing to concerns about the literal clarity of the orders and the competency of both prescribers and nurses who administer the opioids. Policies and practices that construct range orders in a manner that ties a specific dose or route to a pain-intensity rating raise serious concerns about potential harm. This study demonstrates nurses' variable opinions and likely practices related to selecting a dose from within a range and their decisions on how to space doses within a designated time period. For institutions, the findings of this study underscore the importance of conducting a multidisciplinary examination of range order practices and the need to increase efforts to educate prescribers in how to write appropriate range orders and nurses in how to implement them to provide effective and safe analgesia. A closer examination of opioid pharmacokinetics can provide logic to develop a new paradigm where range orders are replaced with titration orders. Clear understanding and communication between staff involved in the medication administration process is essential. Nurses should work together with colleagues in pharmacy and medicine to develop and study the safety of alternative ways of constructing orders that would more clearly guide titration of an opioid to meet a patient's need.Benedetti et al., 1998, ISMP (2002) Source : http://www.medscape.com/viewarticle/580915