Physical Therapy or Advanced Imaging as First Management Strategy Following a New Consultation for Low Back Pain in Primary Care: Associations with Future Health Care Utilization and Charges

Abstract

Low back pain (LBP) imposes a large socioeconomic burden on individuals and health care systems. Average annual direct costs per individual are estimated from $1,500 to $2,000 (Fritz et al. 2012; Martin et al. 2012), with total direct costs for LBP in the United States estimated at over $86 billion in 2005 (Martin et al. 2008). For individuals with LBP seeking health care, a common entry point is primary care (Sundararajan et al. 1998). One of every 17 primary care visits involves LBP (Licciardone 2008). Considering the prevalence of LBP in primary care and its strategic position as the entry point, efforts to improve care increasingly focus on primary care decisions (Scott, Moga, and Harstall 2010; Slater et al. 2012).

A focus of efforts to reduce costs is advanced imaging, particularly magnetic resonance imaging (MRI), for uncomplicated LBP (Chou et al. 2011; Baker et al. 2013) due to high costs and increasing use (Deyo et al. 2009; Lehnert and Bree 2010; Hughes et al. 2011) despite no evidence of benefit to patients in the absence of specific indications (Chou et al. 2009). American College of Radiology guidelines recommend against imaging during the first 6 weeks for uncomplicated LBP with or without radiculopathy (Davis et al. 2009). Indications of complicated LBP that may require imaging include trauma, findings suggestive of neoplasm or infection, and rapidly progressing neurologic deficits (Davis et al. 2009; Chou et al. 2011). Common MRI findings can reduce patients' well‐being (Modic et al. 2005) despite a lack of correlation with symptoms (Jensen et al. 1994; de Schepper et al. 2010). Concern about overuse of imaging is highlighted by inclusion in the “Choosing Wisely” campaign as a top five primary care activity for which change could lead to higher quality and better resource use (Group 2011).

There have been several published efforts to reduce inappropriate LBP imaging in primary care (Rosenthal et al. 2006; Curry and Reed 2011; Georgiou et al. 2011). Effectiveness of these efforts is impacted by several factors, including whether an alternative strategy is available (Roshanov et al. 2013). Patients have expectations around receiving something perceived as beneficial. Breaking an expectation by denying imaging may be unacceptable to patients or providers (Zusman 2013). Consumer research suggests offering an alternative to replace the broken expectation is important to patients (Santa 2013). Physical therapy has been used as an alternative management because it is often viewed as credible and meets patient and provider expectations (Blackmore, Mecklenburg, and Kaplan 2011; Srinivas, Deyo, and Berger 2012), but comparisons of these alternatives (advanced imaging or physical therapy) are lacking. Purposes of this study were to examine patients with a new primary care consultation for uncomplicated LBP who received advanced imaging or physical therapy as the first management strategy. We evaluated associations between the management received and health care utilization and LBP‐related charges in the year following primary care consultation.

Methods

Results

Exactly 3,355 patients with a new LBP‐related primary care consultation were identified, of which 2,893 (86.2 percent) were included (Table 1). Reasons for exclusion are outlined (Figure 1). Among included patients, 841 (29.1 percent) received management outside primary care in the initial 6 weeks. First care received was advanced imaging (n = 385, 45.8 percent) or physical therapy (n = 377, 44.8 percent) in most cases. In the unmatched sample, patients receiving advanced imaging received the image a mean 9.2 days (SD = 9.2) after the index visit, and 112 (29.9 percent) eventually received physical therapy over the 1‐year follow‐up. Almost all imaging was an MRI (n = 363, 94.3 percent). Mean charges for the imaging were $1,306 (SD = $460). Patients receiving physical therapy began treatment a mean 8.4 days (SD = 8.7) after the index visit, receiving a mean 3.8 visits (SD = 3.0). Seventy‐three (19.7 percent) eventually received advanced imaging. Mean physical therapy charges were $504 (SD = $441). There were no instances of beginning physical therapy and advanced imaging on the same date. Patients (n = 79) receiving management outside primary care in the initial 6 weeks (other than advanced imaging or physical therapy) received a physician specialist visit (n = 63, 7.5 percent) or other care (e.g., chiropractic, n = 16, 1.9 percent).

Open in a separate window

Figure 1

Study Sample and Reasons for Exclusion

Baseline differences were identified indicating the advanced imaging group was older, more likely to be male, have more comorbidities, a specific LBP diagnosis, history of previous surgery, and have received opioid or corticosteroid medications at the index visit (Table 1). Propensity matching resulted in 203 pairs with no baseline differences (Table 1).

Odds of each utilization outcome were higher for patients who first received advanced imaging compared with physical therapy (Table 2). Odds ratios attenuated only slightly in the matched sample (Table 2). Mean total 1‐year LBP‐related charges in the unmatched sample were $6,193 (95 percent CI: $4,826, $7,560, median = $2,261) for advanced imaging and $1,467 (95 percent CI: $1,180, $1,755, median = $540) for physical therapy. Mean charges were similar in the matched sample (mean charges = $6,664 [95 percent CI: $5,589, $7,740] median = $2,415) for advanced imaging and $1,871 ([95 percent CI: $1,569, $2,173 $629] median = $629) for physical therapy (mean difference = $4,793 [95 percent CI: $3,676, $5,910]).

Discussion

This study found that if care beyond primary care was provided during the first 6 weeks after a new consultation for uncomplicated LBP, beginning with advanced imaging instead of physical therapy increased the odds of surgery, injections, specialist, and emergency department visits within a year. Propensity‐matched comparisons shifted the odds minimally, indicating much of the difference between advanced imaging and physical therapy was attributable to factors other than patient characteristics. One‐year LBP‐related health care charges were about $4,700 higher with advanced imaging.

Our findings support studies reporting increased risk of additional health care when patients with uncomplicated LBP received an early MRI (Webster and Cifuentes 2010; Graves et al. 2012). It is unlikely these findings only indicate those receiving an early MRI have a more severe condition. For example, a randomized trial comparing early MRI to radiographs also found increased rates of injections and surgery in the MRI group (Jarvik et al. 2003). These results support our finding that patient characteristics do not fully explain the increased risk for future health care utilization with early advanced imaging. Labeling effects may offer a partial explanation. Advanced imaging often “labels” a patient's LBP that might otherwise be viewed as nonspecific and uncomplicated, causing heightened concern in some patients and providers and motivating additional care‐seeking (Kendrick et al. 2001; Modic et al. 2005).

Reasons for early advanced imaging in patients with uncomplicated LBP, in contradiction to guidelines, are likely multifactorial. Patients' level of insistence on pursuing imaging cannot be determined. Although the insurance benefit design did not require our sample to obtain advanced imaging prior to specialist referral, individual specialists may have this requirement. Financial interest in imaging services can motivate utilization (Paxton et al. 2012). Some cases in our sample may have had indications consistent with guidelines for early imaging, such as rapidly progressing neurologic loss. Regardless of motivation, our findings support others in recognizing the early use of advanced imaging increases patients' risk of exposure to more costly, invasive procedures.

We compared early advanced imaging with an alternative of physical therapy. Both unadjusted and adjusted comparisons found early physical therapy was associated with decreased risk of all utilization outcomes and lower LBP‐related charges over 1 year. These findings are consistent with other observational studies reporting reduced risk of subsequent health care utilization and lower costs for patients receiving physical therapy within 2–4 weeks of a new primary care consultation, compared to delayed physical therapy (Fritz et al. 2012; Gellhorn et al. 2012). Physical therapy may avoid the negative consequences of a labeling effect from imaging. When evidence‐based, physical therapy should provide patients with an active approach to LBP, enhancing patients' perceived ability to self‐manage their condition (Breese and French 2012). Evidence indicates that when physical therapy is adherent to guidelines, it can be effective in reducing additional care‐seeking; however, variability in the delivery of evidence‐based care is evident (Fritz et al. 2012; Gellhorn et al. 2012).

The goal of preventing patients with LBP from progressing to invasive and costly procedures has increased focus on initial primary care decisions. Early decisions appear to have important implications for the subsequent course of care (Webster, Verma, and Gatchel 2007; Webster and Cifuentes 2010; Fritz et al. 2012, 2013; Graves et al. 2012, 2014; Webster et al. 2013, 2014). Many patients approach an initial consultation with expectations for aggressive actions that may include imaging (Kendrick et al. 2001; Hoffman et al. 2013). Dismissing this expectation can result in dissatisfaction (Staiger et al. 2005), prompting the need for an alternative (Srinivas, Deyo, and Berger 2012). Some pathways recommend physical therapy (Blackmore, Mecklenburg, and Kaplan 2011; Flynn, Smith, and Chou 2011). Our results support physical therapy as an alternative to advanced imaging for patients or providers with expectations for additional care. The question of which patients should receive care beyond the primary care setting remains an important consideration for future research. Our results indicate that if additional care is sought, physical therapy may be the preferred initial step instead of advanced imaging.

Our results should be considered in light of several limitations. We could not assess all possible confounders. Potentially important variables such as pain intensity or symptom duration were not attainable. We could not measure indirect costs, which are substantial for LBP (Dagenais, Caro, and Haldeman 2008). We did not have patient‐centered outcomes including function or satisfaction. We could not evaluate management with complimentary or alternative providers (e.g., chiropractic, massage, etc.) that may have occurred but was not represented in claims data. The question of which patients need care beyond primary care was not addressed. In our sample about 70 percent of patients did not receive care beyond primary care in the first 6 weeks. Our sample included a single health care system and cannot reflect geographic variability in spine care.

In conclusion, this study found patients newly consulting primary care for uncomplicated LBP who received care beyond primary care had lower risk for future health care and lower charges if they started with physical therapy compared to advanced imaging. Initial management with advanced imaging was associated with increased risk of subsequent utilization and higher LBP‐related charges in the year following primary care consultation in analyses using propensity matching to control confounding. These findings support care models recommending physical therapy instead of advanced imaging for initial management of patients with uncomplicated LBP who want or need care beyond the primary care setting.

Supporting information

Acknowledgments

References