The Role of Expectations in Pain After Dysvascular Lower Extremity Amputation

Abstract

Objective:

The objective was to determine the relationship between pain expectations assessed prior to surgery and satisfaction with pain 4 and 12 months after major dysvascular lower extremity amputation.

Research Method:

The study included a prospective cohort of male (n = 19) veterans experiencing their first lower extremity amputation due to complications of diabetes mellitus or peripheral arterial disease. Measures included presurgical expectations of pain at 4 and 12 months postamputation, actual average pain and satisfaction with pain at 4 and 12 months postamputation, and agreement between expected and actual pain.

Results:

Sixty-eight percent of participants expected to experience no or minimal pain at 4 months; 95% expected to experience no or minimal pain at 12 months. Thirty-two percent and 58% of participants had more pain than they expected at 4 and 12 months, respectively. Participants whose pain expectations were met reported higher satisfaction with their actual level of pain at 12 months postamputation, even after adjusting for current pain levels.

Conclusion:

The results underscore the potential value of fostering realistic expectations about the degree to which amputation may impact average pain.

Introduction

The relevance of expectations in medical settings is apparent, driving practices such as informed consent (Leclercq, Keulers, Scheltinga, Spauwen, & van der Wilt, 2010) and phenomena such as the placebo effect (Crow et al., 1999; Price, Finniss, & Benedetti, 2008). The goal of this paper was to explore the potential impact of expectations on long-term pain among individuals undergoing planned amputation, which may be considered as a means for relieving pain in addition to preserving a limb or restoring mobility (Chin, Teague, McLaren, & Mahoney, 2013; Taylor et al., 2005).

Although this question has been evaluated primarily among individuals undergoing hip and knee replacement surgery, several studies have suggested that expectations around pain following medical interventions can significantly impact satisfaction with pain and with the interventions themselves. Among a sample of individuals undergoing spinal surgeries, having expectations met was the greatest predictor of patient-rated effectiveness of surgery (Mannion, Kampfen, Munzinger, & Kramers-deq Uervain, 2009). In a study of individuals who underwent orthopedic or abdominal surgery, having presurgical pain expectations met (even if the expectation was for moderate to severe pain) was associated with better overall satisfaction with surgery (Svensson, Sjostrom, & Haljamae, 2001). However, many individuals undergoing surgery overestimate the degree to which surgical interventions can improve pain (Mannion, Junge, et al., 2009; Mannion, Kampfen, et al., 2009).

In contrast, other studies have demonstrated that the actual experience of pain is more strongly associated with satisfaction than pain expectations. Among individuals undergoing knee arthroplasty in two separate studies, actual knee pain, not pain expectations, significantly predicted postsurgical satisfaction (Lingard, Sledge, & Learmonth, 2006; Mannion, Kampfen, et al., 2009).

Thus, while it is apparent that expectations may have an impact on outcomes and satisfaction with outcomes, the associations among these variables are poorly understood. The current study extends these ideas to a sample of individuals with recent dysvascular lower extremity amputation (LEA). Our first goal was to describe participants’ presurgical expectations of pain 4 and 12 months postamputation and assess the degree to which these expectations were associated with actual 4- and 12-month pain. Our second goal was to assess if participants whose pain expectations were met were more likely to be satisfied with their pain levels postamputation, and if so, whether this association was maintained after adjusting for actual pain level.

Study Design

This study is part of a larger multisite prospective cohort study of individuals undergoing LEA due to peripheral arterial disease or diabetes mellitus at two VA medical centers and a large academic medical center. The larger study included 87 participants and is described in more detail elsewhere (Norvell, Turner, Williams, Hakimi, & Czerniecki, 2011; Webster et al., 2012). This particular study was conducted with a subset of participants (n = 29) who were recruited prior to amputation surgery; all of these individuals were recruited from VA medical centers. Participants were assessed via in-person interview presurgically and via in-person or telephone interview at 4 and 12 months postsurgery. This study was conducted in accordance with the procedures approved by the human subjects review boards at each study site.

Participants

Subjects were considered eligible if (a) they were age 18 years or older; (b) they were awaiting a first LEA, defined as transmetatarsal level or higher; and (c) the primary cause of amputation was diabetes mellitus or peripheral arterial disease. Of the 29 enrolled presurgically as part of the larger study (who therefore had presurgical expectation data available), 23 (79%) remained enrolled at 12-month follow-up; two of these 29 participants (6.9%) died in the course of the study, and four (13.8%) were lost to follow-up. Nineteen (66%) of 23 study completers had complete pain data at 12 months. Those who died, were lost to follow-up, or who had incomplete pain data were not significantly different from those who remained in the study in any baseline health or demographic factors. Participant characteristics are described in Table 1.

Table 1.

Sociodemographic and Health Data Among Participants

	N= 19
Mean age in years (SD)	63.03 (8.89)
Veteran, n (%)	19 (100%)
Female, n (%)	0 (0%)
Race/ethnicity, n (%)
White	16 (84.2%)
Nonwhite	3 (15.8%)
Amputation level (at 6 weeks)
Below knee (transmetatarsal & transtibial)	18 (94.7%)
Above knee (transfemoral)	1 (5.3%)
Number of pain locations endorsed at 4 months, n (%)
0	6 (31.6%)
1	4 (21.1%)
2	1 (5.3%)
3+	8 (42.1%)
Number of pain locations endorsed at 12 months, n (%)
0	6 (31.6%)
1	3 (15.8%)
2	1 (5.3%)
3 +	9 (47.4%)

Note. Age ranged from 47 to 79 years.

Measures

Pain expectations.

Prior to amputation, all participants were asked, “On average, how much pain do you expect to have 4/12 months after your surgery?” with possible responses ranging from 0 (no pain) to 10 (extreme pain).

Actual pain.

To assess actual pain at 4 and 12 months postamputation, participants were asked, “Do you currently experience any persistent and bothersome pain?”(Ehde et al., 2000) and were given yes/no response options. If individuals responded “no” to the initial pain item, they were given a 0 for their average pain intensity. If participants responded in the affirmative, they were then asked follow-up questions about pain severity in seven different bodily locations (e.g., residual limb, back). For each location noted, participants were asked, “In the past month, on average, how intense was your pain rated on a 0 to 10 scale with 0 being ‘no pain’ and 10 being ‘pain as bad as could be’?”

For individuals who endorsed pain in one or more areas, an average pain intensity score was created for the areas where any pain was endorsed (Ehde et al., 2003; Ehde et al., 2001; Ephraim, Wegener, MacKenzie, Dillingham, & Pezzin, 2005). Numeric rating scores for both pain expectations and actual pain intensity were converted into pain categories using well-established cutpoints, such that scores of 0–4 were classified as no/minimal pain, 5–6 were classified as moderate pain, and 7–10 were classified as severe pain (Fejer, Jordan, & Hartvigsen, 2005; Jensen, Smith, Ehde, & Robinsin, 2001; Jones, Vojir, Hutt, & Fink, 2007). This allowed for meaningful categorization of pain as well as a threshold for evaluating agreement or discrepancy between expectations and actual pain scores.

Agreement between expected and actual pain.

A dichotomous “expectations met” variable was computed by assessing for categorical differences (e.g., minimal, moderate, severe) between presurgical pain expectations and actual 4-month and 12-month pain scores. If participants experienced the expected amount of actual pain or less (e.g., if expected pain fell in the moderate range and actual pain fell in the moderate [5–6] or minimal [0–4] range), pain expectations were considered “met.” If participants experienced more pain than expected postamputation, pain expectations were considered “unmet.”

Satisfaction with pain.

At both 4 and 12 months, participants were asked, “How satisfied are you with your current average level of pain?” Possible values ranged from 0 to 10, where 0 = not at all satisfied and 10 = extremely satisfied.

Data Analyses

To address our first goal, describing presurgical pain expectations and the associations between these expectations and actual pain at 4 and 12 months, we show frequencies of presurgical pain expectations and corresponding pain outcomes, the proportion of participants who experienced more pain than expected, and the proportion of participants whose pain expectations matched their actual pain (e.g., who experienced equivalent or less pain than expected).

To determine if having one’s pain expectations met was associated with satisfaction with pain, forced entry regression analyses were computed with satisfaction with pain at each time point as the dependent variables and the dichotomous pain expectations variable (met vs. unmet expectations) as the independent variable. As several studies have demonstrated differences in both pain expectations and pain perceptions based on gender and age (Campbell, Guy, & Banim, 1999; Gandhi, Davey, & Mahomed, 2009), the analyses were adjusted for age. Because this analysis indicated that having one’s pain expectations met was associated with greater satisfaction with pain, we then conducted forced-entry multiple regression analyses with satisfaction with pain at each time point as the dependent variables and expectations met as the independent variables. This time, analyses were adjusted for age and current actual pain.

Results

Description of Pain Expectations and Pain Outcomes

As shown in Table 1, 42% and 47% of participants reported pain in three or more areas at 4 and 12 months, respectively. Notably, mean pain scores and worst pain scores (i.e., the area in which participants rated the highest level of pain) were significantly correlated at each time point (4- and 12-month Spearman’s ρ = .86 and .89, respectively, both p < .001). A majority of participants (68.4%) expected to experience no or minimal pain at 4 months postamputation; nearly all (94.7%) expected to experience no or minimal pain at 12 months postamputation (see Table 2). Most participants (57.9%) subsequently reported no or minimal actual pain at 4 months, but only 37% reported no to minimal actual pain at 12 months postamputation. Thus, a substantial proportion of participants had unmet pain expectations at 4 months (31.6%), and at 12 months (57.9%).

Table 2.

Expected and Actual Pain at 4 and 12 Months

	Month 4		Month 12
	Expected pain N (%)	Actual pain N (%)	Expected pain N (%)	Actual pain N (%)
No/minimal pain	13 (68.4%)	11 (57.9%)	18 (94.7%)	7 (36.8%)
Moderate pain	4 (21.1%)	4 (21.1%)	0 (0%)	6 (31.6%)
Severe pain	2 (10.5%)	4 (21.1%)	1 (5.3%)	6 (31.6%)

Note. There were four individuals who reported no pain at both 4 months and 12 months postamputation; these individuals were included in all analyses in the “no/minimal” pain category. Expected pain was assessed at presurgical baseline. Actual pain was assessed at 4 and 12 months.

Association Between Expectations and Satisfaction With Pain

At 4 months postamputation, satisfaction with pain was comparable regardless of whether pain expectations were met (M = 6.38; SD = 3.46) or unmet (M = 5.50; SD = 1.22), β = .46; t(17) = 1.19, ns. However, at 12 months, individuals whose pain expectations were unmet reported significantly lower levels of satisfaction with their pain (M = 3.46; SD = 2.84) than individuals whose pain expectations were met (M = 7.33; SD = 2.94), β = .56 t(17) = 2.63, p = .02. Having pain expectations met at 12 months was associated with greater satisfaction with pain, even after adjusting for age and current (12-month) pain, β = .64; t(16) = 2.16, p = .05.

Discussion

These findings suggest that expectations for postamputation pain are often unrealistically optimistic—participants indicated an expectation for decreasing pain over time, yet a majority had their 12-month pain expectations unmet. This is consistent with the literature that indicates that a majority of individuals continue to experience bodily pain in multiple areas after amputation. In a sample with a median time of 7 years since amputation, 52% of participants endorsed persistent and bothersome back pain (Ehde et al., 2001). Similarly, in another study with a median time of 4 years since amputation, 95% of participants endorsed having at least one type of amputation-related pain in the past 4 weeks (with phantom limb, residual limb, and back pain the most commonly reported areas; Ephraim et al., 2005). Congruent with studies of other samples undergoing surgeries (Palazzo et al., 2014), we found that having expectations met was a key determinant of satisfaction with pain 1 year later, even after adjusting for actual pain experienced at that time.

The findings suggest that fostering realistic expectations plays an important role in long-term satisfaction with pain. Identifying factors that influence expectations and developing interventions to foster realistic expectations could potentially be integrated routinely into the surgical decision-making process, perhaps through informed consent, discussion with a rehabilitation team, and formal/informal education.

Study Limitations

There are several limitations to the current study. The small sample size rendered our analyses underpowered. While this lends credibility to the findings that emerged, it limited our ability to examine more complex models that could enrich our understanding of factors contributing to or resulting from expectations. In conjunction with our small sample size, our exclusively male sample also likely biased the findings and limited generalizability (Campbell et al., 1999; Gandhi et al., 2009). Third, while assessing pain with a numeric rating scale is well-established, providing categorical anchors on the measure (to denote minimal/moderate/severe pain) would likely increase its validity as well as clinical applicability (Jensen et al., 2001). Also, the study sample was restricted to participants with some level of ambulatory function prior to amputation and with the cognitive capacity to participate in an interview, which may limit generalizability to other amputee samples, such as individuals undergoing an amputation revision or second amputation, or to individuals with amputation due to trauma. Last, we did not examine factors that might influence expectations. Future research is needed to better understand factors that impact the formulation of pain expectations as well as provide both patients and care providers with a better understanding of what realistic postamputation pain expectations might be.

Conclusions

Understanding the role of expectations in the context of planned amputation may have particular clinical impact, as the surgery marks the beginning of an ongoing partnership between patients and providers. Given that expectations are influenced by and associated with knowledge, hopefulness, self-efficacy, and patient activation, fostering realistic and flexible expectations that are shared by patient and provider is a potential target for ongoing clinical intervention and limb-loss management.

Impact and Implications.

• Though the role of presurgical pain expectations has been prospectively assessed in other medical populations undergoing planned major surgical procedures (e.g., hip and knee replacement surgery), this study extends this line of questioning to individuals undergoing lower extremity amputation.

• This study demonstrates that having one’s pain expectations met is associated with greater satisfaction with pain 1 year postamputation, regardless of actual pain level.

• This study highlights the potential value of developing clinical interventions aimed at fostering realistic expectations around the degree to which amputation may affect pain.