Is referral to a spine surgeon a double-edged sword?

Abstract

Objective

To assess the concerns of adult patients with spine-related complaints during the period between referral to and consultation with a spine surgeon.

Design

Prospective survey.

Setting

Toronto, Ont.

Participants

A total of 338 consecutive, nonemergent patients before consultation with a single spine surgeon over a 5-month period.

Main outcome measures

Patient concerns, effect of referral to a spine surgeon, and effect of waiting to see a spine surgeon.

Results

The issues patients reported to be most concerning were ongoing pain (45.6% rated this as most concerning), loss of function (23.4%), need for surgery (12.1%), and permanence of the condition (9.6%). Regression analysis demonstrated that older age was an independent predictor of increased level of concern regarding pain (P = .01) and disability (P = .04). Forty-seven percent of all patients listed the need for surgery among their top 3 concerns. Mere referral to a spine surgeon (P = .03) was an independent predictor of increased concern regarding the need for surgery. Sex, diagnosis, surgical candidacy, and actual wait time were not predictive of increased concerns. Patients reported family physicians to be their most influential information source regarding spinal conditions.

Conclusion

Timely provision of more specific information regarding the benign and non-surgical nature of most degenerative spinal conditions might substantially reduce patients’ exaggerated concerns regarding the probability of surgery for a considerable number of patients referred to spine surgeons.

Patient satisfaction with clinical encounters is related to the extent to which their concerns have been acknowledged and addressed.^1–4 However, there is often a considerable mismatch between health care professionals’ perceptions of patient concerns and the actual concerns experienced by patients.^3,4 This would seem to be the case in Canada, where patients’ beliefs about treatment and prognosis of low back pain (LBP) are generally pessimistic and contrary to current evidence.⁵ Several evidence-based clinical practice guidelines (CPGs) have noted the typically favourable natural history of LBP and advocate a biopsychosocial approach to professional care and self-management.^6,7 However, persistent or recurrent back symptoms are not uncommon and are not well addressed by CPGs.^6,8–13 Furthermore, many CPGs do not consider patient preference or the logistics of implementing recommendations.¹⁴ Practitioners report that these and other fundamental issues, such as maintaining physician-patient relationships, are reasons for nonadherence to LBP CPGs, which often leads to increased use of health care services including investigations and specialist consultation.^{11–13,15–22} Given the fact that LBP is one of the most common presenting complaints in primary care, this issue represents an important health care challenge.^6,23

In Canada, those referred to spine surgeons can wait many months before consultation.^24–26 Consequently, living in uncertainty of diagnosis, prognosis, and further management might create or perpetuate patient concerns. While there have been surgical studies exploring patient concerns in the perioperative period,^27,28 no studies exist that specifically explore patient concerns that arise during the postreferral–presurgical consultation (PRPC) period. Understanding the factors that contribute to patient concerns during this waiting period could inform strategies to reduce these concerns and enhance self-management.

The primary objective of this study was to determine the nature and extent of patient concerns during the PRPC period. A secondary objective was to assess associations between various factors and reported patient concerns.

METHODS

Consecutive patients were recruited during a 5-month period from a single Canadian academic institution before undergoing consultation with a spine surgeon. This study received institutional research ethics board approval.

Questionnaire development

As this is the first study to address these particular issues, a questionnaire was developed with questions based on the common concerns voiced by patients attending the spinal clinic of the senior author (Y.R.R). This 9-item questionnaire included Likert-scale items and other semiquantitative items using ordinal scales; it also included open-ended questions that allowed for collection of patient comments. We initially assessed the draft questionnaire by having 20 patients complete it; their feedback on the clarity of the questions and suggestions for improvements was incorporated. After several iterations of pilot testing, the questionnaire was deemed to be highly comprehensible. The revised questionnaire was also assessed by an interprofessional team focused exclusively on the care of spinal disorders. This team included a spine surgeon, an acute-care nurse practitioner, inpatient ward nurses, a physiotherapist, an occupational therapist, and a clinical spinal research assistant.

Inclusion criteria

The finalized questionnaire was then administered prospectively to 338 consecutive patients just before consultation (ie, while in the waiting room) with a single spine surgeon over a 5-month period. Owing to regional referral patterns, most patients seen in this clinic have disorders affecting the thoracolumbar spine. This cohort represented approximately two-thirds of all referrals received, as one-third of requested referrals were deemed non-surgical and were declined (ie, were not seen by the spine surgeon) based on the original information provided by the referring physicians and further requested information. All referrals were triaged based on relevant clinical information provided in the referral, and thus variable wait times are represented.

Exclusion criteria

Emergent or urgent surgical cases (less than 10% of referred cases) were excluded. Questionnaires that were more than 50% incomplete were considered invalid. Questionnaires that were completed by patients who appeared not to understand English or the questions asked were also excluded. Lack of comprehension was judged by any 1 of the following 3 criteria: free-text responses that were contradictory to the corresponding multiple-choice responses, multiple-choice responses that contradicted one another, or self-identification (ie, patients indicated that they did not understand). In total 29 (8.6%) out of 338 questionnaires were excluded.

Collection of data

A clinical spinal research assistant, not involved in the medical care of the patients, collected all data.

Baseline demographic data on age and sex were collected. The primary spinal diagnosis was determined and categorized (Table 1) by the spine surgeon at consultation.

Table 1.

Diagnoses: N = 309.

CLINICAL DIAGNOSIS	N (%)
Lumbar* radiculopathy	86 (27.8)
Lumbar* claudication	82 (26.5)
Back pain	57 (18.4)
Cervical radiculopathy	8 (2.6)
Cervical myelopathy	3 (1.0)
Neck pain	10 (3.2)
Coronal deformity	14 (4.5)
Sagittal deformity	6 (1.9)
Tumour or infection	4 (1.3)
Inflammation	6 (1.9)
Multifactorial pain	28 (9.1)
Miscellaneous	5 (1.6)

^*Lumbar is defined as a condition affecting the spine at levels L1 to S1.

Based on clinical evaluation, patients were also prospectively categorized by the spine surgeon (blinded to survey responses) as either surgical candidates or not. Surgical candidates were defined as those whose symptoms, combined with correlative structural abnormalities, would have been amenable to surgery at any point, including those who declined or deferred surgery and those whose symptoms had improved or resolved by the time of consultation.

Consultation wait time was calculated as the number of days between receipt of referral and consultation.

Statistical analysis

Continuous and normally distributed data (age and wait time) were compared using t tests. Categorical variables, such as sex and surgical candidacy, were compared using Yates continuity-corrected χ² test.

Multivariable logistic regression modeling was used to determine associations between certain independent variables and the level of concern patients had regarding ongoing pain, disability, permanence of their conditions, and the possibility of surgery. The independent variables were age, primary spinal diagnosis, sex, being rated as a surgical candidate, and wait time. Patients were dichotomized into low-concern (respondents indicating no concern or little concern) and high-concern (respondents indicating that they were somewhat or very concerned) groups.

Data were analyzed using the SPSS 14.0 statistical software package. All reported P values are 2-tailed, using an α of .05.

RESULTS

We surveyed 338 consecutive patients, resulting in 309 (91.4%) valid questionnaires being returned. In our study cohort, 160 of 338 (47.2%) patients were men. Mean (SD) and median age were 55 (15) and 54 years, respectively, with an interquartile range of 23 years and a range of 18 to 95 years. Approximately three-quarters of patients had diagnoses of lumbar spinal conditions (Table 1). The mean (SD) and median wait times from referral to consultation were 120 (60) and 126 days, respectively, with an interquartile range of 63 days and a range of 3 to 553 days.

Ongoing pain (45.6%), loss of function (23.4%), the perceived need for spine surgery (12.1%), and permanence of symptoms (9.6%) were rated as the top 4 patient concerns about their spinal conditions (Table 2). The degree of concern was high for most respondents (Table 3). Thirty-seven percent of patients were determined by the surgeon to be surgical candidates. Forty-seven percent of all patients listed the need for surgery among their top 3 concerns, with no difference (P = .6) between those who were determined to be surgical candidates (41.2%) and those who were not (35.6%). At 12 and 24 months after completion of the survey, 14.9% and 22.7% of all patients surveyed had actually had or were booked for surgery.

Table 2.

Most concerning aspect of spine condition identified by patients: N = 239.^*

PATIENT CONCERN	TOP CONCERN, N (%)	WITHIN THE TOP 3 CONCERNS, N (%)
Ongoing pain	109 (45.6)	198 (82.8)
Loss of function	56 (23.4)	178 (74.5)
Need for surgery	29 (12.1)	112 (46.9)
Permanence of condition	23 (9.6)	141 (59.0)
Lack of information	11 (4.6)	22 (9.2)
Fear of paralysis	8 (3.3)	31 (13.0)
Fear of cancer	3 (1.3)	9 (3.8)
Other	0 (0.0)	9 (3.8)

^*Not all respondents answered this question.

Table 3.

Degree of patient concern: N = 309.

PATIENT CONCERN	LOW CONCERN,* N (%)	HIGH CONCERN,†N (%)
Ongoing pain	21 (6.8)	281 (90.9)
Loss of function	38 (12.3)	256 (82.8)
Need for surgery	47 (15.2)	219 (70.9)
Permanence of condition	35 (11.3)	234 (75.7)
Lack of information	107 (34.6)	142 (46.0)
Fear of paralysis	151 (48.9)	108 (35.0)
Fear of cancer	201 (65.0)	42 (13.6)
Other	5 (1.6)	22 (7.1)

^*Low concern amalgamates respondents indicating no concern or little concern.

^†High concern amalgamates respondents indicating that they were somewhat concerned or very concerned.

Logistic regression analyses are presented in Table 4. Regression analysis showed that older age was a significant predictor of an increased level of concern regarding ongoing pain (P = .01) and loss of function (P = .04). Being a surgical candidate was not a predictor (P = .61) of having concerns about the possibility of future surgery (Table 4A). Covariates such as age, sex, primary diagnosis, and wait time between referral and consultation were also not significant predictors for increased concern about possibly needing surgery (Table 4A).

Table 4.

Predictors of increased levels of concern: A) Type 3 analyses of effects using a binary logistic model to detect significant predictors of increased level of concern; the probability modeled for each of the main patient concerns was the high-concern group; Wald χ² and P values are shown. B) Odds ratios for positive independent predictors of increased patient concern.

A) PREDICTOR	PATIENT CONCERN
	INCREASED CONCERN ABOUT ONGOING PAIN		INCREASED CONCERN ABOUT LOSS OF FUNCTION		INCREASED CONCERN ABOUT NEED FOR SURGERY		INCREASED CONCERN ABOUT PERMANENCE OF THE CONDITION
	χ2	P	χ2	P	χ2	P	χ2	P
Age	6.37	0.01	4.19	0.04	4.68	0.41	1.02	0.31
Sex	2.10	0.15	1.84	0.18	1.84	0.18	0.09	0.76
Diagnosis	4.18	0.65	10.48	0.10	7.58	0.35	2.15	0.90
Referral effect*	1.97	0.74	8.99	0.06	11.03	0.01	1.64	0.80
Surgical candidate	0.89	0.64	0.35	0.84	0.16	0.61	0.10	0.95
Wait time†	1.02	0.31	1.63	0.20	1.83	0.63	0.38	0.54
Waiting effect‡	3.32	0.34	5.20	0.16	0.40	0.31	4.21	0.24

B) EFFECT	HIGH CONCERN FOR …	ODDS RATIO ESTIMATE	95% WALD CONFIDENCE LIMITS
Age	Ongoing pain	1.06	1.01,1.12
Age	Loss of function	1.04	1.00,1.07
Referral	Need for surgery	3.09	1.27,7.52

^*Referral effect refers to the subjective effect on concerns of being referred to the spine surgeon.

^†Wait time represents the time waited between referral and consultation.

^‡Waiting effect refers to the subjective effect on concerns of waiting for consultation on patients’ levels of concern.

Mere referral to a spine surgeon resulted in increased reported levels of concern (compared with before referral) in 52.9% of patients; when patients were asked why their concerns were increased, the most commonly cited reason was fear of surgery (24.3%). Regression analyses showed that mere referral to a spine surgeon was an independent predictor (P = .03) of developing a high level of concern regarding the possibility of surgery (Table 4A).

Waiting for consultation was associated with an increased level of reported concern in 49.1% of patients; long wait times, increasing pain, and fear of the unknown were the most widely cited reasons in responses to open-ended qualitative questions. Adjusted analyses showed that the actual time waited between referral and consultation was not predictive (P = .63) of an increased level of patient concern (Table 4A).

DISCUSSION

Nearly half (45.6%) of patients waiting for consultation with a spine surgeon reported ongoing pain as their greatest concern. Loss of function (23.4%), perceived need for surgery (12.1%), and permanence of their conditions (9.6%) were also reported as most concerning by respondents. Older age was an independent predictor of high concern regarding pain and loss of function. Of all respondents, 46.9% listed the need for surgery among their top 3 concerns. Having been referred to a spine surgeon was found to be the only independent predictor of increased concern associated with possibly needing surgery.

Research examining patient concerns has focused on eliciting and elucidating the concerns of patients in order to enhance the physician’s ability to better address patient concerns and therefore enhance patient care and satisfaction.^27,29–33 The results of this study provide an increased awareness of the degree and type of concerns that patients experiencing LBP have during and likely before the PRPC period. The unique effect of the surgical referral itself increases concerns about the potential need for surgical intervention and introduces an additional dimension of patient concern that has not been previously considered. In addition, this study also reaffirms that family physicians are reported by their patients to be the most common source of information as well as the most influential regarding level of concern. These findings are in contrast to those of McIntosh and Shaw in the United Kingdom, which suggested patients accessed information about their LBP from a variety of sources.³⁴ Of interest, these authors also found that general practitioners were equivocal about their role as information providers. The findings of our study identify an important opportunity for Canadian primary care physicians to be more influential by proactively discussing and addressing patient-specific concerns at different time points along the course of care. While this might seem simple, physician-patient interactions, particularly for complex problems such as chronic LBP, are far from simple and require physicians to balance their own beliefs and expectations with those of their individual patients and with the external pressures of individual health care environments, all while trying to maintain or build the doctor-patient relationship.^{11,13,16,35,36} In other words, there are many reasons to engage in surgical referral, some of which have more to do with the patient’s illness experience and the family physician’s adherence to patient-centred care and less to do with hard surgical indications.^37,38 A detailed discussion of this topic is outside the scope of this manuscript; however, this study does provide the contextual information to facilitate improved physician-patient communication and knowledge transfer to patients with LBP, who (as an unintended effect of referral to a spine surgeon) often have the incorrect perception that surgery is inevitable.

Several findings of this study warrant discussion about current knowledge that can help address the concerns of patients with chronic LBP. The high degree of concern for ongoing pain and disability is in keeping with the commonly recurrent nature of LBP.^6,8 The finding that these concerns were greater with increasing age is possibly owing to a greater effect (subjective or objective) of spinal disorders in combination with other comorbidities on the elderly. In addition, older patients are more likely to be suffering from spinal stenosis, which tends to be persistent once symptoms occur.^6,39,40 For approximately half of these patients, the symptoms are managed conservatively. However, for those with symptoms of claudication or leg-dominant pain, surgical intervention, when indicated, is associated with a high and sustainable degree of success.^39–42 In contrast, younger patients are more likely to suffer from disk disorders or nonspecific LBP, both of which might be recurrent and serve as substantial sources of patient frustration, but which are typically associated with a favourable natural history and relatively mild symptoms.^6,8,42,43 For those with persistent, severe, uncomplicated, nonmechanical LBP, consideration of and assessment for generalized or complex regional pain syndromes should be pursued.⁶

While the high degree of concern expressed about the need for surgical intervention might be valid from the patient’s perspective (eg, severe symptoms or failed conservative treatment), such concern is unfounded for most patients.⁶ Patients reported a lack of information or knowledge as being one of the least concerning aspects of the their spinal conditions (Table 5). Thus, there appears to be a disconnect between patients’ perceptions of the meaning of their symptoms and whether or not they are candidates for surgery. Given that these individuals waited several months to see a surgeon (plus the duration of pain or treatment before surgical referral) this wait would appear to provide ample time to generate the biopsychosocial debility that tends to accompany chronic conditions.^44,45 Half of the patients surveyed also perceived that waiting for consultation substantially increased their level of concern. Adjusted analyses, however, suggested that the increase in level of concern was independent of waiting effect and actual wait time. In other words, the well-intended or patient-requested surgical referral appears to be a double-edged sword.

Table 5.

Sources of information for patients regarding their spinal conditions

SOURCE	SOURCE OF INFORMATION, N (%)*		MOST INFLUENTIAL ON LEVEL OF CONCERN, N (%)*
	NO	YES
Family physician	49 (15.9)	245 (79.2)	91 (38.1)†
Other orthopedic or neurosurgeon	94 (30.4)	126 (40.8)	53 (22.2)
Other non-surgical specialist	111(35.9)	110 (35.6)	36 (15.1)
Allied health provider	74 (23.9)	173 (56.0)	25 (10.5)
Family	86 (27.8)	128 (41.4)	15 (6.3)
Other (eg, self)	28 (9.1)	14 (4.6)	10 (4.2)
Media	71 (23)	137 (44.4)	6 (2.5)
Friends	90 (29.1)	122 (57.3)	3 (1.3)

^*Percentages do not add to 100% because not all respondents answered all questions.

^†Regarding concerns for the probability of surgery, 36% of respondents in the low-concern group rated family doctors as highly influential, while 58% of the corresponding respondents in the high-concern group rated family physicians as highly influential.

The aim of this survey was to identify patient-specific concerns just before seeing a spine surgeon. The results only reflect patients’ perspectives and do not provide objective information from the referring physicians about the specific reasons for the referrals that were discussed with the patients at the time of referral. Consequently, we are only able to speculate on the possible reasons why almost half of the patients were unnecessarily concerned about an inevitable need for surgery. It is plausible that surgical referral implies (or corroborates patients’ perceptions of) a level of “seriousness” regarding patients’ conditions and generates a fear that they might require surgery, regardless of the intent of the referral. Contrary to this specific patient concern, substantial literature supports the fact that surgery is rarely recommended for LBP and is typically ineffective for chronic nonradicular back pain in the absence of gross instability or deformity such as spondylolisthesis.^39,46–49 In addition, for nonemergent spinal conditions, a trial of non-surgical care is always recommended before spinal surgery is considered.^6,46 Surgery is typically effective compared with failed non-surgical care for well-defined clinical conditions such as radiculopathy, neurogenic claudication, and spondylolisthesis.^{39,40,42,43,46,48} Furthermore, surgery for these conditions is elective and is thus ultimately the decision of the informed patient.

It is most likely that ongoing pain and disability are the main reasons for referral; education and primary management in such cases have usually been exhausted from a physician or patient perspective. In this scenario, specialist referrals are necessary as a means of acquiring both expertise about more specific diagnoses and shared management of the patient’s pain and disability. However, 36% of those reporting low concern about the probability of surgery rated the primary care physician as highly influential, and 58% of those reporting high concern rated the primary care physician as highly influential. This finding identifies an opportunity for the referring physician to address a unique patient concern while such patients are waiting to see spine surgeons. Providing and reinforcing information for patients about the low likelihood of need for surgery, or directing patients to reliable spine-specific websites or other information before and during the PRPC period, might serve to temper some of the concerns found in this study. At a minimum, referring physicians should convey (at the time of referral) firm reassurance that most degenerative spinal conditions are manageable and that most patients with LBP referred for consultation do not require surgery. In addition, as approximately half of patients referred were not likely to benefit from any type of surgical management, we believe that, in light of the often-extensive wait list for surgical consultation, increased concomitant use of non-surgical specialists (eg, physiatrists) might be helpful. This would be particularly important in the clinical scenario of back-dominant pain, as it would likely be a timelier and less anxiety-provoking means of addressing some or all of the ongoing concerns and management needs of patients with chronic LBP. However, the effects of such an approach on patient concerns require further study.

Limitations

A limitation of this study is that it is based on a single report by each participating patient regarding their concerns. In addition, the actual duration of symptoms or the extent of concurrent management were not specifically studied and might represent important confounders. The tertiary care, single-surgeon design also carries the potential for selection bias and might not necessarily be reflective of spinal patients being referred to surgeons regionally or nationally. Furthermore, patients in this cohort were prescreened with a bias toward being more likely to be surgical candidates; the general population of patients with LBP, most of whom do not typically get referred to spine surgeons, are not represented.^8,23 However, most of the diagnoses were representative of common noncomplex spinal disorders (LBP, radiculopathy, and neurogenic claudication). Another limitation is the exclusion of the approximately 10% of patients with urgent or emergent spinal symptoms. This group could represent a subpopulation for which these reported surgical concerns are more likely to be valid; in other words, urgent or emergent neurologic symptoms often require surgery, and thus concerns about inevitable surgery would not be unfounded.

Conclusion

Patients referred to spinal surgeons for nonurgent spinal disorders are understandably most concerned about ongoing pain and loss of function. The kind and degree of patient concerns reported were found to be independent of wait time. Furthermore, a substantial number of patients had unfounded concerns about the need for surgery. Family physicians were found to be the most influential source of information affecting patient concerns and thus are well situated to deliver concern-specific education and counseling before and during the PRPC period.

EDITOR’S KEY POINTS

• In Canada, patients referred for consultation with surgical specialists sometimes wait many months. During this waiting period, patient concerns are unknown and often remain unaddressed. This survey sought to understand the factors that contribute to patient concerns during this wait in order to inform strategies to reduce these concerns, refine referral practices, and enhance self-management strategies.

• Although most patients suffering from benign spinal disorders are not surgical candidates, a large number of participants in this survey were concerned that they were inevitably destined for surgery. Mere referral to a spine surgeon independently increased patient concerns about the need for surgery, regardless of surgical candidacy.

• Patients did not identify lack of information as one of their main concerns; however, the substantial number of patients incorrectly concerned that surgery was inevitable suggests that there is a need to provide, whenever possible, more specific education about the likelihood of needing surgery.

• Patients reported family physicians to be their most influential source of information about spinal conditions. Family physicians are thus well situated to educate patients waiting to see spine surgeons and to reduce their exaggerated concerns before referral.

POINTS DE REPÈRE DU RÉDACTEUR

• Au Canada, les patients dirigés en chirurgie pour une consultation attendent parfois des mois. On ignore ce dont les patients s’inquiètent durant cette période d’attente et leurs inquiétudes demeurent souvent sans réponse. Cette enquête voulait comprendre les facteurs qui contribuent aux inquiétudes des patients durant cette attente afin de trouver des stratégies pour réduire ces inquiétudes, améliorer la façon de demander des consultations et favoriser des stratégies d’autogestion.

• Même si très peu de patients qui présentent des affections bénignes du rachis sont susceptibles d’être opérés, un grand nombre de participants de cette étude s’inquiétaient de devoir nécessairement l’être. Le simple fait de demander une consultation en chirurgie du rachis augmentait de façon indépendante les inquiétudes des patients concernant la nécessité d’une chirurgie, quelle qu’en soit la probabilité.

• Selon les patients, le manque d’information n’était pas la principale préoccupation; toutefois, le nombre élevé de ceux qui s’inquiétaient inutilement de devoir être opérés suggère qu’il y a lieu, chaque fois que c’est possible, de mieux renseigner les patients sur la probabilité d’une telle intervention.

• Selon les patients, c’est le médecin de famille qui est leur source d’information la plus importante au sujet de leur condition rachidienne. Le médecin de famille est donc bien placé pour renseigner les patients qui attendent de voir un chirurgien du rachis et pour apaiser leur craintes excessives durant cette période.