Secondary Logo

Journal Logo

Triage of Spine Surgery Referrals Through a Multidisciplinary Care Pathway: A Value-Based Comparison With Conventional Referral Processes

Wilgenbusch, Chelsea S., BSc(Hon); Wu, Adam S., MD, MSc, FRCSC; Fourney, Daryl R., MD, FRCSC, FACS

doi: 10.1097/BRS.0000000000000574
Roles of Registries and Care Pathways in Defining Value for Spine Surgery
Open
Blog

Study Design. Retrospective medical record review.

Objective. To (1) determine if outpatient referrals for low back pain (LBP) and leg pain triaged through a multidisciplinary spine care pathway (group A) were more likely to be candidates for surgery than conventional physician referrals (group B); (2) compare relevant clinical differences in the 2 groups (e.g., diagnosis, pain scores, level of disability); and (3) compare wait times for magnetic resonance imaging and surgical assessment.

Summary of Background Data. The Saskatchewan Spine Pathway was introduced on the basis of evidence that a co-ordinated, multidisciplinary, and stratified approach to the assessment and management of LBP may improve quality. During early implementation, some physicians began to refer patients to Saskatchewan Spine Pathway clinics, whereas others continued to refer patients directly to the surgeon through the conventional process.

Methods. We retrospectively analyzed consecutive new outpatient referrals for LBP and leg pain, June 1, 2011 through May 30, 2012 for 2 surgeons.

Results. We identified 215 referrals, including 66 (30.7%) in group A and 149 (69.3%) in group B. There was no difference in overall health (mean EuroQol Group 5-Dimension Self-Report Questionnaire visual analogue scale) or lower back-related disability score (Oswestry Disability Index). Group A patients were significantly more likely to be candidates for surgery (59.1% vs. 37.6% for group B; P = 0.0034, χ2 test), had significantly poorer scores for EuroQol Group 5-Dimension Self-Report Questionnaire mobility, a higher proportion of leg dominant pain, and a lower proportion of back dominant pain. Group A patients also had significantly shorter wait times for magnetic resonance imaging and surgical assessment.

Conclusion. A co-ordinated multidisciplinary pathway with a stratified approach to LBP assessment and care provided a greater proportion of surgery candidates than the conventional referral process. The implementation of such processes may allow surgeons to restrict their practices to patients who are more likely to benefit from their services, thereby reducing wait times and potentially reducing costs.

Level of Evidence: 3

In this retrospective study, we showed that triaging elective referrals for low back pain through a multidisciplinary pathway significantly improved the rate of candidacy for surgery. Triaged patients had poorer mobility, more leg dominant pain, and less back dominant pain. The process improved wait times for magnetic resonance imaging and surgical assessment.

From the Division of Neurosurgery, University of Saskatchewan, Royal University Hospital, Saskatoon, Saskatchewan, Canada.

Address correspondence and reprint requests to Daryl R. Fourney, MD, FRCSC, FACS, Division of Neurosurgery, University of Saskatchewan, Royal University Hospital, 103 Hospital Dr, Saskatoon, Saskatchewan, Canada, S7N 0W8; E-mail: daryl.fourney@usask.ca

Acknowledgment date: April 17, 2014. First revision date: June 13, 2014. Second revision date: July 24, 2014. Acceptance date: July 28, 2014.

The manuscript submitted does not contain information about medical device(s)/drug(s).

Supported by AO Spine North America Inc. Analytic support for this work was provided by Spectrum Research Inc., with funding from the AO Spine North America.

Relevant financial activities outside the submitted work: grants.

This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivitives 3.0 License, where it is permissible to download and share the work provided it is properly cited. The work cannot be changed in any way or used commercially.

Low back pain (LBP) has become a major target for quality improvement due to significant variation in clinicians' management, a high risk of chronicity, rising health care resource utilization, and its sizable socioeconomic burden.1 The reasons for continued chronicity and increased cost are related not only to the nature of LBP, but also the current methods of health care delivery in most industrialized nations. Patients face a myriad of health care and provider options, with little or no collaboration among providers, often leading to care that was not co-ordinated, fragmented, and episodic.2

Most adults will experience LBP in their lifetime.3 Patients whose symptoms fail to improve in a timely fashion, or who present with significant neurological involvement, are often referred to a spine surgeon. However, many patients referred for surgical assessment do not require surgery,4–6 and the sheer volume of these referrals has resulted in long waiting lists to consult spine surgeons in Canada.5–8 As a result, many patients who are candidates for surgery endure long waits to consult the surgeon, whereas those who are not good “surgical candidates” may experience delays in the initiation of optimal nonsurgical care.5 Worse outcomes have been associated with these delays, including less pain relief and reduced likelihood of return to work.9,10

In 2010, the Province of Saskatchewan, Canada launched a new model of care on the basis of evidence that a stratified approach to assessment and management of LBP and interprofessional care may be helpful. The goals of the Saskatchewan Spine Pathway (SSP) were to improve quality for LBP care by encouraging guidelines-concordant evidence-based primary care while reducing wait times for appropriate magnetic resonance imaging (MRI) and surgical referral. The SSP uses a multidisciplinary triage process to quickly identify which patients are likely surgical candidates and which patients are likely nonsurgical candidates, allowing for more efficient triaging of surgical referrals and more appropriate utilization of MRI for patients in the first group, and more timely initiation of appropriate nonoperative care for patients in the second group, reducing delays associated with obtaining unnecessary imaging and/or surgical referral in these cases. This is done partly through triage clinics staffed by specialized SSP physiotherapists, but the pathway also promotes online training and live courses to improve guidelines-concordant primary care for LBP including referral patterns. The details are published elsewhere.1

During the early implementation of the SSP, some physicians began to refer patients to the SSP clinics, whereas others continued to refer patients directly to the surgeon through the conventional process. The objectives of this early implementation study were to: (1) determine if surgery referrals from the SSP clinics were more likely to be candidates for surgery than conventional referrals; (2) determine relevant clinical differences in the groups (e.g., diagnosis, pain scores, level of disability); and (3) examine wait times for MRI and surgeon assessment.

Back to Top | Article Outline

MATERIALS AND METHODS

This study was approved by the University of Saskatchewan Biomedical Ethics Board (Bio no. 13–37). A retrospective medical record review was performed of all new elective outpatient referrals for LBP and leg pain seen by 2 neurosurgeons in the Division of Neurosurgery, University of Saskatchewan (D.R.F., A.S.W.), from June 1, 2011 to June 1, 2012. Six referrals were excluded because of clinical red flags (fever/chills, symptoms of cauda equina syndrome, tumor, etc.) and were treated as emergency referrals.

Patients were stratified into 2 groups on the basis of the method of referral to surgery: group A were referred by the SSP clinic, whereas group B were referred directly by primary care physicians through conventional referral processes.

We defined a surgical candidate as a patient offered surgery by the surgeon. Surgery was offered if the surgeon judged that surgery was indicated to correct a putative anatomical source of pain or diminished neurological function (or both), on the basis of the correlation between clinical presentation and imaging findings. In determining candidacy, the surgeon also took into account the response (or lack thereof) to nonsurgical therapies, the time course and natural history of the particular condition, and the medical fitness of the patient. General indications for surgery included disc herniation, spinal stenosis, spondylolisthesis, adjacent segment disease, and/or symptomatic mechanical disc collapse. After shared decision-making, most patients signed consent for surgery, except for 3 of 66 (4.5%) in group A and 7 of 149 (4.7%) in group B. However, for the purposes of this study, these nonconsenting patients were still considered surgical candidates.

A detailed description of the SSP is beyond the scope of this article, but has been published elsewhere.1 It was designed to improve quality patient-focused care for patients with LBP by encouraging guidelines-concordant primary care. The SSP has 3 key features: (1) educational resources for physicians and patients, (2) specialized clinics, and (3) outcomes research to monitor quality. The training course can be found online (available at: www.spinepathwaysk.ca). Training has been completed by the majority of Saskatchewan primary care physicians, chiropractors, and nurse practitioners.11 Patient information, including self-care strategies and shared decision-making modules for spinal stenosis and herniated disc can be found online (available at: http://www.health.gov.sk.ca/back-pain). The SSP classification (SSPc) is a clinical classification of symptoms and signs for mechanical low back and leg pain modified from the work of Hall et al,12,13 to stratify care. Evidence-based treatment algorithms (self-care, mechanical therapy, medications) follow from the SSPc and can be found online (available at: http://www.health.gov.sk.ca/provider-spine-info).

Patient flow through the SSP is shown in Figure 1. Patients with significant red flags are referred directly to the spine surgeon on call. Patients without red flags who fail to improve by the timelines specified in the algorithm for the particular SSPc pattern may be referred for additional investigation to SSP clinics in the cities of Saskatoon and Regina. These clinics, staffed by specially trained physiotherapists, reassess the SSPc diagnosis as well as the treatments and patient education that have been provided. The goal of the SSP clinic is not to provide comprehensive LBP care, but rather to help educate and direct patients to receive further mechanical therapy, imaging, and/or referral to a spine surgeon when indicated. Priority referrals for MRI and surgery, facilitated by the SSP, are one of the several strategies incorporated to promote primary care compliance.

Figure 1

Figure 1

Data were obtained from the SSP clinic and surgeon charts. Data included patient age, sex, dates of initial referral, imaging, and clinic assessments, whether or not patients were offered surgery by the surgeon, and date of surgery. Visual analogue scale (VAS) back and leg scores, Oswestry Disability Index, and EuroQol Group 5-Dimension Self-Report Questionnaire (EQ-5D) scores were routinely collected at the time of surgeon assessment.

The SSPc patterns of mechanical pain are: back dominant pain aggravated by flexion (pattern 1); back dominant pain aggravated by extension but not increased with flexion (pattern 2); leg dominant pain that is constant, associated with positive neurological findings and aggravated by back movement (pattern 3); leg dominant pain that is intermittent, aggravated by activity in extension and relieved by rest in flexion (pattern 4). Pattern 1 is divided into “fast responders” (increased pain on flexion and relief with unloaded passive lumbar extension) and “slow responders” (increased pain on flexion and extension), but for the purposes of this study we did not analyze these subgroups. In determining the SSPc, leg dominant patterns (patterns 3 or 4) took precedence. Back-dominant patterns are mutually exclusive; however, leg dominant patterns may coexist. If the patient was determined to have elements of patterns 3 and 4, then pattern 3 took precedence. Because of the natural history of LBP, the SSPc for an individual patient may change over time. We collected the SSPc at the time of surgical consultation.

Back dominance or leg dominance was determined in 2 ways: by SSPc and by VAS pain scores. For SSPc, patterns 1 or 2 are back dominant and patterns 3 or 4 are leg dominant. For VAS categorization, patients with VAS back score more than VAS leg score were considered “VAS back dominant” and those with VAS back score less than VAS leg score were considered “VAS leg dominant.” Patients in the “VAS no dominance” category had equal VAS back and VAS leg scores. We also analyzed differences in mean VAS back score and mean VAS leg score in groups A and B.

Wait time for surgeon assessment was calculated as the number days from the initial referral until the surgeon saw the patient. In group A patients, this included both the time to be seen by the SSP clinic and subsequently the surgeon.

To reduce the potential for bias, wait time for MRI was determined only for patients who did not have MRI completed at the time of the initial referral to the SSP clinic (group A) or the surgeon (group B). In Saskatchewan, primary care access to MRI is limited, but may be ordered by the SSP clinic or the surgeon. MRI was often requested by the surgeon prior to his assessment, depending on the clinical information provided in the referral. The wait time for MRI was calculated as the number of days from submitting the request until completion of the study.

Back to Top | Article Outline

Statistical Analysis

χ2 tests and the Student t test were performed as indicated to assess differences in groups A and B. Two-sample proportions test was done to compare SSPc and back versus leg dominance between group A and group B patients. To examine the difference in wait times for surgeon assessment and MRI, the Wilcoxon 2-sample test for nonparametric data were performed and the median is reported. A confidence interval of 95% was used to establish significance and all results are reported to a 95% confidence interval.

Back to Top | Article Outline

RESULTS

Objective 1

We identified 215 consecutive elective surgeon referrals for low back and leg pain, including 66 (30.7%) in group A and 149 (69.3%) in group B. However, SSP clinic referrals were significantly more likely to be candidates for surgery (group A 59.1%, group B 37.6%; P = 0.003, χ2 test).

Back to Top | Article Outline

Objective 2

Table 1 outlines the clinical characteristics of both groups. There was no statistically significant difference in age or sex. There was no difference in overall health (mean EQ-5D VAS), lower back-related disability score (Oswestry Disability Index) or mean VAS back/leg pain (Table 1).

TABLE 1

TABLE 1

Group A referrals were significantly more likely to have VAS leg dominant pain and less likely to have VAS back dominant pain compared with group B referrals; however, VAS was unable to determine back versus leg dominance in more than 30% of patients in both groups (Table 1). Conversely, an SSPc pattern diagnosis was determined for all patients. By this method of classification, leg dominant pain was more prevalent than back dominant pain in both groups; however, leg dominant pain was proportionally more common among group A (75.8%) compared with group B (55.0%) referrals (P = 0.0004, 2-sample proportions test) and back pain was more common among group B (45.0%) compared with group A (24.2%) referrals (P = 0.004, 2-sample proportions test).

With regard to the various domains of the EQ-5D, there were no differences in the categories of self-care, usual activities, pain/discomfort, and anxiety/depression. However, group A patients had significantly poorer mobility scores (P = 0.02, χ2 test). We further analyzed EQ-5D results by dichotomizing each of the health states into “no problems” (i.e., level 1) “problems” (i.e., levels 2 or 3), thereby changing the profile into frequencies of reported problems. This was done because numbers of reported level 3 problems was very low in some groups. As in the initial analysis, only the only EQ-5D domain that was significantly different between group A and group B was mobility: the odds of having mobility problems was significantly lower in group B than group A (odds ratio = 0.32, 95% confidence interval: 0.12–0.87).

Back to Top | Article Outline

Objective 3

Patients referred through the multidisciplinary pathway had significantly shorter wait times (Table 2). Mean MRI wait times for group A were less than half as long as group B (27.4 ± 23.0 vs. 63.5 ± 42.4 d, P < 0.0001 Wilcoxon 2-sample test). The wait for assessment by the surgeon was also significantly shorter (69.1 ± 73.7 in group A vs. 129.6 ± 109.1 d in group B, P < 0.0001 Wilcoxon 2-sample test).

TABLE 2

TABLE 2

Back to Top | Article Outline

DISCUSSION

The recent trend to encourage spine team redesign for LBP similar to teams addressing other chronic conditions has produced some evidence that systematic care processes may improve value. In a health maintenance organization setting in the Southwestern United States, Klein et al14 showed reductions in rates of imaging, specialist referral and spine surgery after primary care education and establishment of a spine clinic to triage surgical referrals. In Denmark, the rate of lumbar disc surgical procedures was significantly reduced through the centralized co-ordination of spine clinics incorporating guidelines-concordant medical care, education, physical conditioning, and cognitive behavioral concepts.15 A Michigan insurance company recently showed an almost 30% decrease in surgical rates while maintaining patient satisfaction by mandating the patient visit a physiatrist-led spine clinic for biopsychosocial evaluation and education before it would authorize a surgical consultation.16

Some work has been done to explore the role of nonphysician clinicians as screeners, including physiotherapists, chiropractors, and nurse practitioners.6,17 In a survey of Canadian spine surgeons, Busse et al6 recently showed that a majority of respondents (75.3%) agreed that they would be comfortable not assessing patients with LBP-related complaints referred to their practice if the indications for surgery were ruled out by nonphysician clinicians.

Kindrachuk and Fourney11 previously showed that 71.3% of surgeon referrals were avoided and MRI utilization was reduced by 52.9%, simply by redirecting patients waiting for spine surgery referrals to physiotherapy-led SSP clinics for screening. However, this study is the first to compare conventional referrals with those screened by the SSP. Although conventional referrals were significantly less likely to be candidates for spine surgery (37.6% vs. 59.1%), the absolute difference is less striking than we had expected. This is likely due to the SSP's concurrent initiative to educate primary care providers regarding guidelines-concordant care, including appropriate spine surgery referrals. From March 2010 through the end of the study period, approximately 50% of the primary care physicians in Saskatchewan took the SSP course, and this may account for the relatively high rate of surgical candidates in group B. Historical billing data showed that prior to the implementation of the SSP, only about 15% of new spine referrals were candidates for surgery,11 and similar rates have been published by spine surgeons in other Canadian centers.4,5

Although overall health (EQ-5D VAS), disability (Oswestry Disability Index), and mean VAS pain scores for back or leg were not significantly different between groups, those referred through the SSP were significantly less likely to have back dominant pain and more likely to have leg dominant pain. Thus, it is not surprising that patients referred from the SSP had poorer EQ-5D mobility scores. The SSPc was able to determine back versus leg dominance in all cases, whereas the VAS score for back versus leg could not determine dominance in more than 30%.

With regard to the individual patterns of the SSPc, there were significant differences between groups A and B only for patterns 1 and 4 (Table 1). There were too few patients with pattern 2 pain to calculate any difference. Pattern 1 is most commonly due to nonspecific mechanical type LBP and surgery is seldom indicated, whereas pattern 4 is often neurogenic claudication due to lumbar stenosis, where surgery has been shown to be superior to nonsurgical treatment.18 The SSP appropriately triaged significantly fewer pattern 1 and more pattern 4 patients to surgeon assessment, than conventional physician referrals. Pattern 3 is sciatica most often associated with lumbar disc herniation; the proportion of these patients was not significantly different between groups, suggesting that both the SSP and primary care were equally proficient at properly triaging these cases for surgical referral.

This study has several weaknesses. It is a retrospective review of only 2 spine surgeons. Cost of care was not calculated. A significant source of potential bias was that the surgeon was not blinded to the SSPc or the fact that the patient had gone through the SSP without successful symptom relief. Likewise, if the patient had not gone through the SSP, he may be more biased to recommend nonoperative treatment first. It is also important to point out that we looked at surgical candidacy, not the actual rate of surgery. We also did not assess subsequent patient referrals: for example, whereas nonoperative care may have been recommended as the initial treatment, surgery may have been offered at a later time. The indications for surgery are also likely to differ significantly among surgeons in different practice settings, limiting the generalizability of our results.

One of the aims of the SSP is to improve quality by reducing wait times for MRI and surgeon assessment. The majority of patients with low back and leg pain do not require either MRI or assessment by a spine surgeon, and benefit the most from timely initiation of appropriate guideline-concordant primary care. Stratification of referrals by a standardized assessment process and classification of symptoms allows the surgeon to focus practice on patients who are most likely to benefit from surgery, whereas patients who are less likely to benefit from surgery are steered to nonsurgical care in a timely fashion early on in the assessment process. The diversion of this large pool of patients who do not require MRI or spine surgeon referral in turn reduces the load of unnecessary demand on these services. The process used by the SSP to determine candidacy for surgical referral have been published elsewhere.1,11

Although we found shorter wait time for MRI and surgeon assessment in the SSP group, it is important to point out this may not so much be an “effect” of the pathway, because it is intrinsic to the architecture of the pathway: to encourage implementation in primary care practice, appropriate patients with SSP are prioritized for MRI and surgery clinic appointments.

We recognize that to assess the value of the SSP to the health care system as a whole, we need to assess all patients with LBP (surgical and nonsurgical) managed through both the SSP system and conventional processes with respect to clinical presentation, outcomes, utilization (imaging and surgery), and cost of care. A prospective study of patients who underwent spine surgery is ongoing and we are working to fund a larger study to encompass nonsurgical outcomes. However, it is currently not possible to capture meaningful clinical data regarding nonsurgical patients who are not referred to SSP clinics. As a result, our studies to date have focused on assessing the value of streamlining surgical cases through the multidisciplinary SSP versus the conventional referral process. However, more effective population health with less variation in practice may be achievable for LBP using an SSP-like approach to encourage guidelines-concordant care.

The SSP was specifically designed to address the unique set of circumstances within a Canadian health care system, and it remains to be seen if the principles we have applied may have value in other jurisdictions. Adoption of LBP pathways in the United States may increase as Accountable Care Organizations become responsible for care across settings and more entities adopt the Institute for Healthcare Improvement Triple Aim as part of health care reform.

Back to Top | Article Outline

CONCLUSION

The SSP, which is a co-ordinated multidisciplinary pathway with a stratified approach to LBP assessment and care, selected a higher proportion of candidates for surgery than the conventional referral process. The difference in absolute surgical yield between groups may have been more striking if it were not for the fact that SSP clinics were introduced concurrently with educational courses targeting primary care to encourage more appropriate surgical referrals. The stratified approach using the SSPc selected patients with more leg dominant pain and fewer patients with back dominant conditions, likely accounting for the difference in surgery rates between the 2 groups.

Although the SSP benefits the surgeon by triaging more appropriate referrals, the system is designed to be patient-focused while addressing the needs of the patient, the payer, and society. This study was not designed to assess cost of care or outcomes between conventional care and pathway care. However, it suggests that implementation of the SSP may allow surgeons to focus their practice more on patients who are more likely to benefit from their services, potentially reducing wait times.

Back to Top | Article Outline

Key Points

  • The SSP selected a higher proportion of appropriate candidates for surgery than conventional physician referrals.
  • There were no differences in overall pain or LBP-related disability between groups; however, the SSP selected more patients with leg dominant pain and fewer patients with back dominant conditions, likely accounting for the difference in surgical candidacy rate between the 2 groups.
  • Implementation of the SSP may allow surgeons to focus their practice more on patients who are more likely to benefit from their services, thereby reducing wait times.
Back to Top | Article Outline

References

1. Fourney DR, Dettori JR, Hall H, et al. A systematic review of clinical pathways for lower back pain and introduction of the Saskatchewan Spine Pathway. Spine (Phila Pa 1976) 2011;36:S164–71.
2. Foster NE, Hartvigsen J, Croft PR. Taking responsibility for the early assessment and treatment of patients with musculoskeletal pain: a review and critical analysis. Arthritis Res Ther 2012;14:205.
3. Hoy D, Bain C, Williams G, et al. A systematic review of the global prevalence of low back pain. Arthritis Rheum 2012;64:2028–37.
4. Deis N, Findlay JM. Appropriateness of lumbar spine referrals to a neurosurgical service. Can J Neurol Sci 2010;37:843–8.
5. Fourney DR. Waiting lists for lumbar spine referrals in Canada: what is the solution? Can J Neurol Sci 2010;37:719–20.
6. Busse JW, Riva JJ, Nash JV, et al. Surgeon attitudes toward nonphysician screening of low back or low back-related leg pain patients referred for surgical assessment: a survey of Canadian spine surgeons. Spine (Phila Pa 1976) 2013;38:E402–8.
7. Hurlbert RJ, Mobbs R, Teo C. Access to spine care: a tale of two cities. Can J Neurol Sci 2008;35:308–13.
8. Braybrooke J, Ahn H, Gallant A, et al. The impact of surgical wait time on patient-based outcomes in posterior lumbar spinal surgery. Eur Spine J 2007;16:1832–9.
9. Folman Y, Shabat S, Catz A, et al. Late results of surgery for herniated lumbar disk as related to duration of preoperative symptoms and type of herniation. Surg Neurol 2008;70:398–401; discussion 401–392.
10. Nygaard OP, Kloster R, Solberg T. Duration of leg pain as a predictor of outcome after surgery for lumbar disc herniation: a prospective cohort study with 1-year follow up. J Neurosurg 2000;92:131–4.
11. Kindrachuk DR, Fourney DR. Spine surgery referrals redirected through a multidisciplinary care pathway: effects of nonsurgeon triage including MRI utilization. J Neurosurg Spine 2014;20:87–92.
12. Hall H, McIntosh G, Boyle C. Effectiveness of a low back pain classification system. Spine J 2009;9:648–57.
13. Wilson L, Hall H, McIntosh G, et al. Intertester reliability of a low back pain classification system. Spine (Phila Pa 1976) 1999;24:248–54.
14. Klein BJ, Radecki RT, Foris MP, et al. Bridging the gap between science and practice in managing low back pain. A comprehensive spine care system in a health maintenance organization setting. Spine (Phila Pa 1976) 2000;25:738–40.
15. Rasmussen C, Nielsen GL, Hansen VK, et al. Rates of lumbar disc surgery before and after implementation of multidisciplinary nonsurgical spine clinics. Spine (Phila Pa 1976) 2005;30:2469–73.
16. Fox J, Haig AJ, Todey B, et al. The effect of required physiatrist consultation on surgery rates for back pain. Spine (Phila Pa 1976) 2013;38:E178–84.
17. Sarro A, Rampersaud YR, Lewis S. Nurse practitioner-led surgical spine consultation clinic. J Adv Nurs 2010;66:2671–6.
18. Weinstein JN, Tosteson TD, Lurie JD, et al. Surgical versus nonsurgical therapy for lumbar spinal stenosis. N Engl J Med 2008;358:794–810.
Keywords:

referral; spine surgery; clinical care pathway; health resource; quality improvement; utilization

© 2014 by Lippincott Williams & Wilkins