Pain Management Decision Making Among Long-Term Care Physicians and Nurses

Abstract

The purpose of this study is to explore attitudes and beliefs that affect decisions about prescribing and administering pain medications in older adults who live in long-term care (LTC), with a particular emphasis on those with cognitive impairment. At each of the four participating LTC facilities, data were gathered from three separate groups of health care professionals: physicians, registered nurses, and registered practical nurses. Based on grounded theory, a model was developed that highlighted critical decision points for nurses and physicians regarding pain management. The major themes that emerged from the data concerned pain assessment (lack of recognition of pain, uncertainty about the accuracy of pain assessment and diagnosis) and treatment (reluctance to use opioids, working to individualize pain treatments, issues relating to physician trust of the nurse on prescribing patterns). These findings may facilitate the development of innovative approaches to pain management in LTC settings.

Very little research has been conducted in the area of pain management among older adults who live in long-term care (LTC) settings and have cognitive impairment. These older adults are particularly vulnerable to unrelieved pain because of declining verbal ability coupled with erroneous beliefs among their health care providers about pain. To date, research has focused on innovative methods of assessing pain in older adults (see Hadjistavropoulos, 2005), but issues and barriers around the treatment of pain using medications have not been addressed adequately. In particular, the factors that affect decision making relating to prescribing and administering pain medications from a multidisciplinary perspective need to be studied.

Pain Management in LTC

In Canada, less than a third of residents who live in LTC are cognitively intact, and approximately 40% of LTC residents in Ontario have extreme cognitive impairment (Proctor & Hirdes, 2001). Pain is a frequent problem for these older residents, affecting more than 50% of those who reside in nursing homes (Kaasalainen & Crook, 2004; Proctor & Hirdes, 2001; Roy & Thomas, 1986; Teno, Kabumoto, Wetle, Roy, & Mor, 2004). Teno et al. (2004) found that LTC residents who sometimes experienced excruciating pain tended to be younger and seriously ill with functional decline and weight loss. Williams, Zimmerman, Sloane, and Reed (2005) found that pain was more commonly reported in for-profit LTC facilities.

Despite these alarming findings, pain continues to be poorly assessed. Williams et al. (2005) found that more than one-third of LTC residents (34.7%) had no formal pain assessment conducted. Over the past decade, significant advances in pain assessment have been made through the use of both verbal and behavioral indices (Fuchs-Lacelle & Hadjistavropoulos, 2004; Hadjistavropoulos, 2005; Kaasalainen & Crook, 2004). Research related to the development and evaluation of innovative and effective ways of treating pain among frail older adults, however, is limited.

Beliefs and attitudes of professional care providers play a significant role in the management of pain in older adults. Beliefs often stem from common misconceptions about pain and aging, such as the following: (a) Pain is a normal consequence of aging; (b) aging results in clinically significant reductions in pain sensitivity; (c) patients with dementia cannot feel pain; (d) the potential side effects of opioids in older adults make them too dangerous to use in older adults; or (e) the absence of a report of pain means the absence of pain (Middleton, Knezacek, Robinson, Hartley, & Kaasalainen, 1997; Pasero, Reed, & McCaffery, 1999). Fortunately, these claims are being challenged by researchers (Ferrell, 1996; Kaasalainen & Molloy, 2001).

Beliefs and attitudes about pain management in older adults continue to infiltrate clinical practice in LTC and lead to inaccurate assessments and unnecessary morbidity for an older person. These barriers to inadequate assessment appear to be even more problematic among older adults with cognitive impairment. Despite their high rates of pain, older adults with cognitive impairments appear to be at great risk for undertreatment of pain (Kaasalainen et al., 1998; McCaffery, Ferrell, & Pasero, 2000). For example, Kaasalainen et al. (1998) found that residents with cognitive impairments were prescribed significantly less scheduled medication and received significantly less pain medication than cognitively intact residents. Physicians who work in LTC are more removed from patient care and have less frequent patient-physician visits as opposed to acute care settings. As such, physicians rely more on the staff nurses for updates on pain assessment and for evaluation of the effectiveness of treatment strategies. Clearly, optimal communication must occur among health care providers in LTC, including registered practical nurses (RPN), registered nurses (RN), and physicians to facilitate better pain management.

Despite physicians’ efforts to prescribe pro re nata (prn; i.e., given as needed) pain medications, nurses may sometimes fail to identify pain, and consequently, patients may go without therapeutic pain medications. Katsma and Souza (2000) found that nurses were more likely to believe and document a grimacing patient’s self-report of pain than that of a smiling patient. They also found that older nurses with more experience were less likely to believe or document their patient’s self-report of pain than younger nurses with fewer years of experience. This finding is particularly relevant as nurses who work in LTC tend to be older (Finnick, Crosby, & Ventura, 1992).

McCaffery, Ferrell, and Pasero (2000) surveyed 400 nurses across the United States and found a tendency for nurses’ personal opinions about the patients’ pain, rather than their recorded assessments, to influence choice of opioid dose and to contribute to undertreatment of pain. Similarly, Kaasalainen et al. (1998) found that although RNs assessed residents as having pain, they did not always give pain medication; that is, nurses’ pain ratings were not significantly correlated with the administration of pain medications. Even if clinicians assess and record pain accurately, they may be planning pain management based on their own beliefs rather than patient presentation. A greater focus on nurses’ and physicians’ decision making around prescribing and administering pain medication is needed to inform the development and evaluation of innovative interventions aimed at improving pain management for older adults who live in LTC settings.

Purpose

The purpose of this study was to explore the decision-making process of pain management of physicians and nurses and how their attitudes and beliefs about pain affect their decisions about prescribing and administering pain medications among older adults who live in LTC. We placed a special emphasis on persons with cognitive impairment. The following research questions were addressed:

1. How do LTC physicians, RNs, and RPNs describe their current practices related to pain management among residents of LTC settings?

2. How do LTC physicians, RNs, and RPNs make decisions about giving pain medications to residents with or without cognitive impairment?

3. What are the beliefs and attitudes of LTC physicians, RNs and RPNs that affect decisions about giving pain medications to LTC residents?

4. What would improve current practices related to pain management in LTC?

Method

A qualitative design guided by grounded-theory methodology was used to answer the research questions (Strauss & Corbin, 1998). Ethics approval was obtained from a university-based ethics review board and from each of the participating LTC facilities. Data were gathered from staff at four LTC facilities in a region within southwestern Ontario, Canada. These four LTC facilities were chosen because they (a) allowed RPNs to administer medications and (b) had more than 10 RPNs or 4 RNs working on a day shift that was needed for the focus groups. At each facility, data were gathered from three separate groups of health care professionals: physicians, RNs, and RPNs.

Letters of invitation were mailed to all physicians who were on staff at each of these LTC facilities. The physicians were then contacted via telephone approximately 2 weeks after the letters had been mailed to obtain verbal consent to participate. Nine out of 45 physicians (20%) agreed to participate in a 1-hr interview at a location that was most convenient to the physician. A trained interviewer conducted face-to-face interviews with each of the physicians using a semistructured interview guide (available on request from the first author). The interview questions focused on (a) current knowledge and practices related to pain management in older adults; (b) barriers that are encountered during patient care, including beliefs and attitudes about pain in older adults; (c) the influence of cognitive impairment on pain management strategies; and (d) perceived needs or gaps in knowledge and practice related to pain management. Interview questions were adjusted and modified to pursue various theoretical issues that arose as the interviews progressed. All interviews were audio recorded and transcribed verbatim.

The RNs and RPNs from four LTC institutions were recruited to participate in focus groups held on-site at each institution for the convenience of the participants. Focus groups would have been more difficult to coordinate for the physicians because of their sporadic and individual visits to LTC facilities.

Brown (1999) suggests using a homogenous sample for focus groups to ensure a level of compatibility among participants and reduce the risk of power imbalances and lack of respect for differing opinions. Therefore, groups of RPNs and RNs were interviewed separately to create a more comfortable environment for them to discuss their own perspectives or share issues related to the RN and RPN working relationship in the context of pain management. A trained moderator facilitated focus group discussions. Also, a second moderator was present to observe the nonverbal communication in the group and was responsible for taking field notes during the session (Brown, 1999). Four RN and four RPN focus groups were conducted, which had a total of 24 and 33 participants, respectively. The median number of participants for the focus groups was 7 (minimum: 5; maximum: 10), which fell within the recommended range of approximately 5 to 10 participants per group (Brown, 1999; Morgan, 1997).

For both the physician interviews and nursing focus groups, theoretical sampling was used; that is, data gathering was driven by the concepts derived from the evolving theory (Strauss & Corbin, 1998). Data were collected from multiple sources (i.e., physicians, RNs, RPNs) and methods (i.e., focus groups, interviews) to maximize data variation.

More than half of the RN group (53%) were older than 46 years, 96% were female, and 27% had worked in LTC for more than 10 years. Almost three-quarters of the RPNs (74%) were aged 46 and older, 97% were female, and 64% had worked in LTC for more than 10 years. The mean age of the physician group was 52 years; 67% were male, and all of them had an independent family practice with hospital privileges. These sample characteristics are similar to those found in other studies that have been conducted in long-term care; specifically, a large majority of nurses were female and tended to be older, and physicians working in long-term care were family practitioners (Schmidt & Svarstad, 2002; Walker, McGeer, Simor, Armstrong-Evans, & Loeb, 2000).

Data Management and Analysis

The qualitative software program, N-Vivo (QSR International, 2002), was used to manage the data. The team used suggestions provided by Strauss and Corbin (1998) to free up our thinking about pain management in LTC from possible biases, assumptions, patterns of thinking, and knowledge gained from experience. These strategies included questioning the data to think about possible categories and their dimensions and systematically comparing our categories to the data and our expectations. Data were analyzed separately by two individuals for all interviews to minimize any idiosyncratic biases; they met regularly throughout the analysis phase to discuss progress and findings. In a collaborative fashion, the research team met to review progress; we utilized an iterative process to question and discuss emerging findings.

Data analysis began with open coding, in which concepts or categories were identified and their properties and dimensions were discovered in the data (Strauss & Corbin, 1998). Theoretical comparisons were made as we moved to building theory. As the data were analyzed, changes were made to the interview guide and codebook to reflect emerging themes. Axial coding involved the process of relating categories to their subcategories and of examining how categories crosscut and linked. Selective coding—the final step in the analytic process—involved integrating and refining the emerging theory. In this stage, the major thematic pieces were woven together to create a model or flowchart that highlighted time sequences, major decision points and events, and factors affecting the decisions related to the management of pain in LTC residents. To examine focus group interaction data, we used a descriptive analysis technique so we could understand the common experiences of participants, the statements that evoked conflict or collegiality among participants, and how focus group interactions were built on one another (Duggleby, 2005; Stevens, 1996).

We used a number of strategies to maintain what Strauss and Corbin (1998) refer to as an objective stance. We obtained multiple viewpoints of the phenomenon by including those of physicians, RNs, and RPNs across four LTC facilities. After each physician interview and focus group, informants were mailed a two-page summary of key findings and asked to review and provide feedback about the investigators’ interpretation of the interview data.

Findings

Generally, participants felt that pain management in LTC is problematic and that it requires more attention among clinicians, beginning with appropriate assessment strategies and moving on to effective pain treatment modalities. Although interview and focus group questions centered on pain management, it was clear that the treatment of pain was closely linked to pain assessment. From the data, a model was developed that highlighted major decision points for nurses and physicians around pain management in LTC. These major decision points are represented in rectangular boxes in Figure 1 within an iterative loop.

Figure 1.

Conceptual Overview of Major Themes and Decision Points of Physicians and Nurses Around Pain Management in LTC

Note: LTC = long-term care.

This iterative loop of pain management reportedly began with assessing pain. Physicians and nurses felt that pain assessment was critical and that it set the stage for how pain would be managed for residents in LTC. After an assessment of pain was undertaken, the next decision resulted in either prescribing pain medications, nonpharmacological interventions, or nothing; this action was said to be the responsibility of a physician, although some nonpharmacological interventions could be initiated by nurses without a physician’s order (e.g., massage, distraction, positioning). Next, if treatment modalities were prescribed to manage pain, the nurses then had to make some decisions (e.g., route, timing, dose) about giving the pain treatments, especially if they were ordered on a prn basis. After the medications were given, both the physician and nurse were involved in evaluating the side effects of the medications. Side effects were often evaluated simultaneously with the reassessment of residents’ pain. Nurses and physicians reported the importance of input from family members and other health care providers (e.g., health care aides, pharmacists, physiotherapists, occupational therapists, social workers), often in a consultative role to facilitate the process of pain management.

This model of decision making for pain management was used as the background. The major themes that emerged from the data were woven into the model in terms of their point of influence into the decision-making process. The major themes of the study data are displayed in the context of the model as cloud-shaped forms in Figure 1 and included (a) lack of awareness of residents’ pain, (b) uncertainty of accuracy of pain assessments, (c) reluctance of health care providers to use opioids, (d) working to individualize pain treatments, and (e) issues relating to physician trust of the nurse on prescribing patterns. These themes are discussed individually below.

Lack of Awareness of Residents’ Pain

Nurses and physicians spoke about the widespread concern that resident pain is not being recognized by health care providers who work in LTC. Some participants addressed the issue that pain assessment is complicated by the pervasive negative myths about pain and aging that seem to continue to linger in society. For example, an RPN stated that “some residents may not feel that their pain is important” or some residents may feel that their pain is just a normal part of aging and are reluctant to complain because they have a fear that they might be labeled as trouble makers or complainers. Unless residents choose or are able to communicate their pain to health care providers, their pain may go unnoticed.

Nurses and physicians felt that residents with dementia were at great risk of not having their pain assessed adequately because of their inability to express pain verbally. According to one physician,

Most of your alert residents would get routine Tylenol, not that we were ignoring the ones who aren’t [alert], but it’s true you notice that you don’t give pain medication routinely to people who are cognitively impaired because they are not the ones that are complaining to you.

Another reason for the underassessment of pain was suggested by both nurses and physicians: Some assume that the absence of a pain report means the absence of pain. One physician commented that “people in long-term care don’t complain of pain; you have to go looking for it.” Likewise, a nurse stated, “Because people think: ‘Well she’s in bed, she’s resting, she’s comfortable.’ Well guess what, she’s in bed nothing to do, looking there at the ceiling, and thinking about her pain.”

Uncertainty About the Accuracy of Pain Assessment and Diagnosis

Even if residents did have their pain assessed, there remained some lingering uncertainty about the accuracy of the pain diagnosis which appeared to affect the way pain was managed. The impact of dementia on the perceived accuracy of pain assessments was a significant concern for both physicians and nurses. One physician said, “I can talk about it [pain] but it is damn hard to assess, you know.” The major factors that contributed to their uncertainty were (a) the inadequacy of currently used tools in practice (if any were being used) and (b) their inability to discriminate between pain and other problems such as delirium and dementia. This lack of confidence in the accuracy of their pain assessments was reflected in the way they treated residents’ pain.

I think a lot of times, through faults of our own [nurses] and faults of the physicians, that we see people with dementia as not having pain, and they are not being put on anything for it. It would be interesting to see if you give somebody pain medication if their behaviors change.

Nurses and physicians disclosed their uncertainty around the cause of resident behaviors during assessments, and they were unsure if these behaviors were pain related or because of some other underlying cause, such as delirium, dementia, or another condition. In most cases, they would treat the cause of these behaviors as being something other than pain. “If it was pain, unfortunately it’s the last thing we look at with the dementia clients,” commented one physician.

As a consequence, physicians would try to treat challenging behaviors with a pragmatic or experimental approach. “You resort to pain meds because you’re not sure they are acting out because it is pain or not.” Without a definitive diagnosis of pain, physicians are challenged in their treatment options and are constantly questioning their decisions about chosen treatments:

We are going to try them on Tylenol, and if Tylenol doesn’t work, then we are going to go up to an opioid … so then the question becomes how high do we keep increasing it to the point where we think, well maybe it’s not actually pain, maybe this is part of the dementing process, not pain.

Reluctance of Health Care Providers to Use Opioids

Physicians and nurses shared their feelings about their reluctance to use opioids in older adults who live in LTC. If opioids were prescribed or given, it usually happened after all other nonopioid treatments had been tried and were unsuccessful. Physicians stated that they try to avoid using opioids as much as possible unless there is “a pretty good reason why we need to put them on an opioid.” Instead, physicians would use some alternative strategies to treat pain in residents who live in LTC:

But if you’ve got back pain or your neck hurts, well you know you don’t need all these fancy pills. Less is better. In the nursing home they [the residents] don’t put out much effort … they don’t have to walk that much … they don’t have to really rely on those things. I think a mind-altering drug or a minor tranquilizer like Serax would be more useful than a narcotic.

Physicians and nurses were even more uncomfortable with using opioids with residents who had dementia. One RPN stated, “We really don’t use morphine or opioids for dementia patients.” This reluctance to use opioids in residents with dementia was in part because of the shortcomings of pain assessments in these residents. Nurses and physicians were not confident in their assessments of pain because of the declining verbal abilities of dementia residents and thus were not willing to take the risks to treat pain with stronger medications, such as opioids. For example, one physician stated that “dementia patients are not able to complain so we really have to foresee, assess properly before we put them on opioids.”

Another important theme that emerged from the data was the influence of a palliative care designation in terms of how pain was managed using opioid medications. Both physicians and nurses were more agreeable to using opioids in residents who were deemed palliative. One physician stated, “The only people in my practice that I prescribe narcotics to would be people who are palliative.” An RN presented another perspective around pain management and palliative care: “We tend to focus too much on pain control for palliation as opposed to just everyday clients. Certainly nobody wants to die in pain, but nobody wants to live in pain either.”

A fear of residents becoming addicted to opioids also limited use of opioids. Nurses and physicians discussed their apprehensions about using opioids in that they did not want to produce high tolerance levels in LTC residents. They also stated that this concern was present among family members. One physician commented, “The one that is abused a lot is Tylenol #2 because they get to take it, they get to like it and you know every month they need more Tylenol #2.”

Working to Individualize Pain Treatments

In light of these challenges, both nurses and physicians discussed the need to work toward an individualized approach to pain treatments. They commented on the variability of residents’ tolerance levels to pain and their desired goals related to pain management; not all residents needed or wanted to achieve a pain-free state. “Pain is a funny thing, and you are not necessarily going to get rid of every little bit of everybody’s pain,” commented one physician. Another physician described this viewpoint clearly:

Depends on the person. Some patients say, “I want this pain gone, I want it gone right now and you better do it.” Then there are other patients who say, “I just want the edge off it and there’s times when it’s bad, if you can help me through those times I can cope with the rest.”You know you work with them differently.

Often physicians use trial and error and a “start low and go slow” approach to treating pain with medications. That is, “you start with a particular dose and then you very carefully increase the dose and are watching for side effects.” In this manner, the side effects of pain medications could be carefully assessed simultaneously with the effectiveness of the medications in reducing pain.

Managing the side effects of pain medications was of utmost concern for both nurses and physicians and was constantly at the forefront of their thinking when treating residents’ pain using medications. This concern was in part because of the extra risks of using medications in the older population. As one physician stated: “They [older adults] get all kinds of side effects to the drug because of their age, and they can’t handle it as well. You’ve got to use trial and error.”

Nurses were primarily concerned with managing constipation and nausea for residents who were taking pain medications. Most nurses stated that they would ask the physician to order a laxative or antiemetic at the same time that a pain medication was ordered. Other areas of concern for nurses around managing pain medications included (a) the need to watch for increased lethargy in residents, which could increase the risk for falls, especially for those with dementia, and (b) the potential for residents to develop liver disease from chronic use of acetaminophen. These concerns were echoed among the physicians who were also concerned about drug toxicity that could result from prescribing pain medications for older adults.

Physicians and nurses were also concerned about polypharmacy and associated risks for older residents in LTC. Comments from nurses indicated that they constantly worked toward minimizing the number of medications that residents were taking for fear of residents developing serious problems in the future. Newly admitted residents were particularly concerning. Nurses stated that residents would often arrive at a LTC setting with bags full of old medications, not knowing which ones had been previously discontinued. Nurses seemed quite pleased when they were able to reduce the number of medications for residents, including pain medications, with the mindset that “less is more.”

Both physicians and nurses felt that they needed to negotiate with the resident a balance between the desired pain level and the extent of the side effects of pain medications while addressing the following question: Where is the resident most comfortable? “Because theoretically we [physicians] can increase the medications as high as we want them to but sometimes they [residents] get the side effects to the point where they’re not willing to go any higher.”

The concern of overmedicating residents was expressed repeatedly by both physicians and nurses; this subtheme reached data saturation quickly during the interviews. A primary goal for physicians when treating pain through medications was “to give them a better quality of life but also a safe quality of life; something that is not going to cause something else that is worse than the original pain.” Another physician remarked, “You can have pain control without being zonked.”

Influence of Physician Trust of the Nurse on Prescribing Patterns

The nature of the physician-nurse relationship was another important theme that repeatedly emerged from the data related to how physicians managed pain for residents. Because most physicians have limited patient contact and on-site visits in LTC, they need to rely on nurses to assess pain, monitor side effects of pain medications, and evaluate the effectiveness of pain medications.

If it is a facility I am not familiar with, I would want a lot of detail about the nature of the pain, why [the nurse] is calling me, what’s the source of the pain. If they’ve done an assessment, what are their vital signs, what has been used before, and why are we getting to the point now that we need a different pain medicine? So I rely on the nurses’ skills quite a bit to give me a clear indication of why we need to up or change a pain medicine.

Communication and trust between the physician and nurse was discussed as being a critical aspect of effective working relationships, which ultimately influenced resident care. Physicians were concerned with the level of clinical skill of nurses which was determined largely by the manner and quality of nurses’ reports that were given to physicians. That is, nurses who offered a clear and well-thought-out conversation about a resident’s status tended to be viewed as trustworthy and competent by the physician. On the other hand, those nurses who were unable to do so would be regarded as untrustworthy.

When I phone a nurse, and the nurse can’t pronounce the name of the drug or can’t spell the name of the drug or doesn’t know what kind of drug that it is that I am talking about … then that does not inspire confidence … if the nurse comes on and says, “gee, I don’t usually work on this floor and I don’t know this person,” obviously I am not terribly confident of that nurse in that situation.

If the physician had little trust or respect for the nurse’s clinical skill and decision-making ability, the physician was more reluctant to prescribe pain medications other than acetaminophen. Some physicians spoke of the specific characteristics of the nurse and how these characteristics either enhanced or limited the amount of trust that the physician held for that particular nurse. Physicians come to know nurses whom they can trust based on previous experiences of working with them.

If you get to know the nurses that respect the residents, whatever their degree of dementia, that care about them, and that are clearly working in the best interest of the resident and not just in their own best interest … those are the ones that I have a lot of trust in. If I know the person at the facility where I work … if it’s one of the nurses there that I know, I would be able to come to the conclusion quicker, based on what I know about their clinical skills.

Some nurses were cognizant of this issue as well and understood the physicians’ viewpoint:

These doctors are very busy … a lot of times I find that they are very rushed for their time when they do come here. And I think a good rapport with them does make a difference. If they trust your assessment, they are a lot more eager to prescribe something at that point based on your opinion.

Barriers and Facilitators

The participants identified many barriers and facilitators in optimal pain management for residents in LTC (Table 1). For both physicians and nurses, lack of knowledge emerged as a very strong barrier, particularly around ways of assessing and treating pain using both nonpharmacological and pharmacological interventions. Ineffective communication and lack of time were two other pervasive barriers that seemed to exist across all health care provider groups. For the nurses, competing demands and the nature of their heavy workload in providing resident care challenged them on a daily basis.

RPN1: I mean, one thing when you are doing your med pass, it could be a very long med pass, so you might not be getting that pain medication to the resident right when they need it. Or you might be a half an hour or so and that could be a long time for somebody in pain by the time we get to that person. Particularly in the type of units that we’re on, they are on split units, so if you have people in pain on both sides … because of the workload you are not going to get to them like you should you know.

RPN2: Yeah, it has a lot to do with the workload.

RPN3: I know. We can’t get to all those people at the same time, it’s not possible, we are only one person.

Table 1.

Barriers and Facilitators to Optimal Pain Management in LTC

Barriers	Facilitators
Lack of time/high workload Lack of compensation for physicians for time spent in LTC	Knowledge for all health care providers around pain management
Language barriers for residents	Good communication with patients, family, and other health care providers
Physician availability to LTC nurses	Interdisciplinary collaboration
Conflict among health care providers	Documentation of pain assessments and treatments
Lack of continuity of care across shifts and settings	Resources (e.g., clinics, coworkers, medical specialists, conferences)
Lack of research in the area of pain management in LTC Poor knowledge transfer to health care providers who work in LTC

Note: LTC = long-term care.

Discussion

The findings of this study contribute to our understanding of the decision-making processes for physicians and nurses around pain management in LTC. In particular, these findings help illuminate the manner in which beliefs and attitudes related to pain influence current practices in LTC, the link between pain assessment and treatment, the lack of recognition of pain (including the complicating factor of pain recognition in patients with dementia or delirium), uncertainty about the accuracy of pain assessment and diagnosis, reluctance to use opioids, increased use of opioids in patients labeled as palliative, fear of addiction, balancing pain relief with medication adverse events, concern about overmedication, and issues related to trust and communication among physicians and nurses.

Pain management practices were clearly described by physicians and nurses, beginning with the assessment of pain and moving toward treating it using both nonpharmacological and pharmacological strategies. The barriers and challenges to assessing pain often precluded optimal pain treatments as health care providers were not confident in their choices of treatment strategies, including pain medications. Physicians and nurses reported being careful in their use of pain medications, particularly around opioid use, which often limited their ability to manage pain effectively. A common goal for both nurses and physicians was to work toward an individualized approach that attempted to balance both the risks and benefits of using pain medications. In this way, attempts were made to manage pain for residents without causing more symptoms or side effects that were worse than the original pain. These findings highlight the need to develop supportive mechanisms for health care providers to help them make therapeutic decisions related to managing pain within an individualized approach to care.

The impact of dementia for residents appears to have a significant influence on physicians’ and nurses’ decision making; it is a common thread across all aspects of pain management. These residents represent an extremely vulnerable population in LTC, and their pain is often overlooked unintentionally because of the challenges and barriers to obtaining accurate pain assessments. This contention is well supported by the literature; many studies have shown that pain is underassessed and undertreated in residents with dementia (Feldt, Ryden, & Miles, 1998; Horgas & Tsai, 1998; Kaasalainen et al., 1998; McCaffery et al., 2000; Sengstaken & King, 1993). Without appropriate assessment tools and sufficient knowledge, staff are left empty handed to manage pain for this population who have limited verbal abilities. Efforts are desperately needed to develop and evaluate standardized approaches to pain assessment so that pain can be regularly and properly assessed and ultimately alleviated in this vulnerable population.

Various professional beliefs and attitudes and the way they affected medication decisions were highlighted in our findings. Specifically, nurses and physicians expressed a multitude of concerns around opioid use in residents who live in LTC. Physicians and nurses were reluctant to use opioids, and some physicians acknowledged that they would not prescribe opioids to residents in LTC under any condition. The reasons for underutilization of opioids in older adults described in this study are consistent with other findings in the literature and include poor quality of pain assessments and concerns over polypharmacy, opiophobia, addiction, and other adverse effects (Ardery, Herr, Hannon, & Titler, 2003; Auret & Schug, 2005; McCaffery et al., 2000) These findings highlight the need for future education of health care providers to counteract such pervasive beliefs about pain and aging.

A number of facilitators that would help improve current pain management practices were identified by participants. The need for effective communication and interdisciplinary collaboration was repeatedly described by both physicians and nurses. The nurse-physician relationship was key to providing effective pain management to residents, particularly in regard to the use of pain medications. Schmidt and Svarstad (2002) also found that the quality of pharmacological management was positively associated with the quality of nurse-physician communication. However, the nature of providing comprehensive and consistent care in LTC is fraught with challenges. Registered nurses are becoming further removed from direct care of residents, and RPNs are taking more responsibility for important resident-care decisions and are often the ones who communicate to physicians about resident-care concerns over pain management. Thus, interventions to improve pain management should facilitate sound decision making by nurses and nurture effective communication between nurses and physicians. Perhaps the emerging roles of the nurse practitioner and the clinical pharmacist in LTC settings and electronic health record technologies to provide alerts and treatment algorithms will assist in facilitating nurses’ decision making and nurse-physician communication.

Another important finding was the perceived lack of knowledge articulated by both physicians and nurses around pain assessment strategies and managing pain using medications. The need for more knowledgeable health care providers related to pain management in LTC has been well documented in the literature (Fox, Solomon, Raina, & Jadad, 2004; Martin, Williams, Hadjistavropoulos, Hadjistavropoulos, & MacClean, 2005). Future work is needed to implement and evaluate innovative ways to educate staff with the goal of producing a change in practice.

The barriers and facilitators that were identified by the study participants are similar to those identified in previous work in pain management in LTC (Fox et al., 2004; Martin et al., 2005). Martin et al. (2005), who did not focus specifically on medication decisions, reached similar conclusions related to barriers and facilitators in pain management in LTC using a qualitative study design. They examined the barriers to effective pain assessment and management among older adults with and without dementia from the perspective of LTC nurses, LTC managers, older adults, and informal caregivers of older adults with dementia. Their findings revealed several barriers to optimal pain management in LTC, including accessibility of appropriate services, shortcomings of self-report methods of pain assessment, communication barriers between patients and health professionals, insufficient specialized training in pain assessment and management, and questions about the adequacy of prescribed medications.

In our study, nurses were faced with immense time pressures and appeared to be working against time to meet the demands of their heavy workload. Finding time for pain management seemed insurmountable. They spoke of their frustrations in trying to organize their time so that they could manage resident pain better, which evoked lively interactive discussion within some focus group interviews. During group discussions, they seemed to validate their frustrations with each other and try to find common ground in terms of the quality of care that they provided to LTC residents. Bowers, Lauring, and Jacobson (2001) studied how nurses manage time and work in LTC and found that time pressures forced nurses to forego the “should-do” work to complete the “must-do” work, often resulting in adverse consequences for nurses and residents. In light of our findings, pain management is perceived as a should-do task for nurses. Clearly, efforts are desperately needed to move practice forward and make pain management a must-do obligation for both LTC nurses and physicians.

Recommendations

Based on our study findings, we propose a list of recommendations intended to overcome barriers and improve pain management for residents in LTC:

1. Implement staff development and educational activities related to pain management (e.g., use of appropriate assessment tools, use of opioids) and provide incentives for staff to attend these activities.

2. Assess pain regularly and use an interdisciplinary pain flowsheet or protocol to facilitate routine pain assessments, documentation, and communication of residents’ pain levels among health care providers.

3. Use pain assessment tools based on residents’verbal abilities. For example, the Pain Assessment Checklist for Seniors with Limited Ability to Communicate (Fuchs-Lacelle et al., 2003) has been recommended in a recent systematic review as an appropriate tool to use for assessing pain in older adults with severe dementia (Zwakhalen, Hamers, Abu-Saad, & Berger, 2006).

4. Consider residents with end-stage diseases (e.g., renal failure, congestive heart failure, extreme dementia) as palliative so that more attention may be given to managing their pain.

5. Develop a template to assist nurses when communicating with physicians over the phone about a resident’s pain status. This template should include a detailed and completed pain assessment that addresses (a) onset and history of pain; (b) location, quality, and intensity of pain; (c) precipitating and alleviating factors; and (d) pain medications currently used, their effectiveness, and side effects noted.

6. Individualize pain treatments to balance risks and benefits (pain control vs. side effects) using a “start low and go slow” approach to using medications and negotiate goals for pain management with resident and family member if possible.

7. Utilize other health care providers to help manage resident pain (e.g., health care aides, nurse practitioners, pharmacists, physiotherapists, social workers) to optimize pain treatments using pharmacological and nonpharmacological interventions.

8. Make pain management a priority to discuss during shift reports and care conferences and document these discussions to share among the health care team.

“We still haven’t reached that goal where people are dying pain free … never mind dying pain free, they should be living pain free.”—Nurse participant