Financial impact associated with implementation of the low back pain clinical practice guideline in outpatient physical therapist practice at a large academic medical center

ABSTRACT

Introduction

Low back pain (LBP) is the top health condition requiring rehabilitation in the United States. The financial burden of managing LBP is also amongst the highest in the United States. Clinical practice guidelines (CPGs) provide management recommendations and have the potential to lower health costs. Limited evidence exists on the impact of CPG implementation on downstream medical costs.

Objective

To examine the impact of CPG implementation in physical therapist (PT) practice on direct and downstream costs for patients with LBP.

Methods

A retrospective observational study examined billing data from 270 patients with LBP who were treated at multiple sites within one large academic medical center by PTs who participated in a multifaceted CPG implementation program. Costs were analyzed for direct PT services, downstream medical services, and PT utilization from September 2017 to March 2018 (pre-implementation group) and compared with costs from June 2018 to December 2018 (post-implementation group).

Results

Direct PT costs were significantly lower post-implementation than pre-implementation mean: $2,863 USD (SD: $1,968) vs. $3,459 USD (SD: $2,838), p = .05, 95% CI [11, 1182]. All downstream costs were lower post-implementation with statistically significant lower costs found in downstream imaging: p = .04, 95% CI [32, 1,905]; pharmacy: p = .03, 95% CI [70, 1,217]; surgery: p = .03, 95% CI [446, 9,152], and “other”: p = .02, 95% CI [627, 7,920].

Conclusion

Implementing the LBP CPG in outpatient PT practice can have a positive impact on lowering downstream costs and the potential to increase the value of PT services.

KEYWORDS:

• Low back pain
• practice guideline
• cost and cost analysis

Introduction

Low back pain (LBP) is the number one cause of years lived with disability (YLD) in the United States, accounting for 10.68% of the total YLDs (Institute for Health Metrics and Evaluation, 2019). The prevalence of LBP in the United States is estimated at 12.7 thousand cases per one hundred thousand population (Institute for Health Metrics and Evaluation, 2022). More than seven million American adults cite low LBP as the cause of their disability and activity limitations (Ma, Chan, and Carruthers, 2014). LBP also carries a significant financial burden (Dieleman et al., 2016; Kim et al., 2019; Owens, Hegmann, Thiese, and Phillips, 2019) with an estimated $87.6 billion USD being spent on managing LBP between 1996 and 2013 by individuals, insurance companies, and the United States government (Dieleman et al., 2016). Many of the costs associated with management of LBP are spent on surgery, imaging, pharmacology (including injections), and conservative treatment options, including services provided by physical therapists (Kim et al., 2019; Owens, Hegmann, Thiese, and Phillips, 2019).

Many treatment approaches are not based on evidence, resulting in unwarranted variations in the care provided (Steinberg et al., 2011). As one example, ineffective interventions for LBP, such as therapeutic ultrasound, which has been shown to have little to no impact on short-term pain, function, or well-being, add to the cost of care without improving clinical outcomes (Ebadi et al., 2020). Advanced imaging, surgery, and opioid prescriptions also add unnecessary costs for most cases of LBP without resulting in improved clinical outcomes (Flynn, Smith, and Chou, 2011; Gidwani et al., 2016). Additionally, they create potential harm by exposing patients to radiation, medical complications associated with surgery, and the development of pain catastrophizing and fear avoidance behaviors (Flynn, Smith, and Chou, 2011). Management strategies that are proven effective and cost-efficient are needed.

One strategy to overcoming unwarranted variation in clinical practice and lowering health care spending is the development and implementation of clinical practice guidelines (CPGs), which are statements and recommendations based on systematic reviews of literature, which delineate a course of care or treatment pathway for a specific health condition (Steinberg et al., 2011). Patient care that aligns with CPG recommendations, also known as CPG adherent care, is associated with lower healthcare costs than non-adherent or guideline discordant care (Castillo et al., 2020). When initiated in a timely fashion, evidence-informed physical therapist management (including the use of a CPG) has been shown to diminish disability related to LBP and lower costs of surgery, injections, and emergency department visits (Childs et al., 2015; Fritz, Brennan, and Hunter, 2015; Fritz et al., 2015). When physical therapist services alone do not fully resolve a patient’s LBP, additional health care services may be provided. The charges for additional services such as pharmaceutical management, radiological services, emergency department visits, physician, orthopedic or neurologic specialist visits that are incurred after an episode of physical therapist services are considered downstream costs. To help reduce downstream costs and costs related to excessive provider services, research is needed to determine how best to achieve this. Examination of the impact of a CPG is one method currently available to assist in this effort.

The United States outspends other high-income countries on prescription medications, diagnostic imagining, and hospital and physician visits while remaining at the bottom when it comes to life expectancy at birth, infant mortality, persons living with two or more chronic conditions, and obesity rates (Squires and Anderson, 2015). Therefore, research into cost-effective management strategies for expensive conditions such as LBP are a priority in the United States as the cost of health care continues to rise while the outcomes of population health remain low. Research focused on the timing of physical therapist services has increased the body of evidence on both patient outcomes and direct cost of physical therapist services (Childs et al., 2015; Fritz et al., 2015). However, a gap exists in the literature on the impact of CPG implementation in physical therapist practice on downstream medical costs. The primary aim of this study was to examine the impact of implementing a LBP CPG in outpatient physical therapist practice for patients with LBP on direct and downstream costs. It was hypothesized that direct and downstream costs associated with the management of patients with LBP by outpatient physical therapists would change after the implementation of the LBP CPG compared with management provided prior to implementation.

Methods

Study design and setting

A retrospective observational analysis was performed on patients with LBP who received treatment at one of eleven orthopedic outpatient PT clinics within a large academic medical center between September 2017 and December 2018. The study was approved by the Institutional Review Board (protocol number 2019E0074). The medical center where the data was obtained has a population estimate for the year 2019 of 898,553, of which 51% are female; 59% are White, 29% Black, 6% Hispanic or Latino, and 6% Asian (US Census Bureau QuickFacts: Columbus OH, https://www.census.gov/quickfacts/columbuscityohio). Ten percent of individuals living in the same region as the medical center aged 65 and younger do not have health insurance and 9% identify as a person with a disability (US Census Bureau QuickFacts: Columbus OH, https://www.census.gov/quickfacts/columbuscityohio).

Downstream medical costs, direct PT service costs, and PT utilization were measured before and after the implementation of the LBP CPG titled “Low Back Pain Clinical Practice Guidelines Linked to the International Classification of Functioning, Disability and Health” funded by the Orthopedic Section of the American Physical Therapy Association (Delitto et al., 2012). Detailed downstream cost items included services for pharmaceutical management; imaging services (i.e. radiological, electromyogram, ultrasound, magnetic resonance technology, computerized tomography of head, body, or general); emergency department services; surgery; other rehabilitation (i.e. occupational therapy, speech language therapy, respiratory therapy, and cardiology); and “other” services (i.e. clinic fees, medical surgical supplies and devices, and other non-defined services). Medical costs that were incurred prior and concurrent to physical therapist services, as well as direct PT costs were also examined. All cost items were measured within a six-month time frame, occurring either before or after the implementation of the LBP CPG.

Physical therapist training

A multifaceted implementation strategy was undertaken and included dissemination of the most current LBP CPG (Delitto et al., 2012); formation of a Low Back Outcome team; educational sessions prioritizing a standardized treatment-based classification system (Alrwaily et al, 2016); LBP CPG; and pain science resources (Bantick et al., 2002; Butera, Fox, and George, 2016; Lehman, 2017; Martin and Moseley, 2015; Moseley and Arntz, 2007; Mosely and Butler, 2015; Ogunlana, Odole, Adejumo, and Odunaiya, 2015). The clinicians were not directed to incorporate a specific adaptation from these resources but encouraged to incorporate the knowledge gained in its entirety into their clinical practice. The Low Back Outcome team consisted of 23 physical therapists working at one of the 11 outpatient clinics within the academic medical center. The delivery of educational interventions for clinicians occurred through a multimodal approach which included didactic education through an online platform, recorded versions of the didactic education for on-demand future review, and electronic handouts of pertinent CPG materials. Educational interventions were developed using the Physical therapist-driven Education for Actionable Knowledge translation (PEAK) program (Jk and Mickan, 2014).

During the online educational session for the team, members were polled in real time through an online voting platform to assess their application of the CPG through five clinical vignettes. These questions specifically focused on clinical decision-making as it related to patients’ signs and symptoms. Additionally, questions were incorporated into educational content for team member reflection and to assess their understanding of the Low Back Pain treatment pathway and how their day-to-day clinical workflow would change.

Data collection

Data were requested from the academic medical center’s Information Warehouse and included billing records for all patients who met the inclusion criteria, which included physical therapist or medical diagnosis of low back pain classified via International Classification of Diseases, 10^th revision (ICD-10) codes of M54.10; M54.17; M54.5; M54.16; M54.12; M54.40; M54.41; M54.42; M99.03; M99.04; M99.05; M51.26; M51.24; M40.30; and men and women 18–80 years of age. Exclusion criteria included patients who had back surgery in previous 12 months.

Data were collected for two time periods. The first period, labeled pre-implementation, occurred between September 2017 and March 2018, and represented 6 months prior to the implementation of the LBP CPG training. The second period, labeled post-implementation, occurred between June 2018 and December 2018, and represented 6 months into CPG implementation (i.e. 3 months were provided for the implementation process). Individuals with LBP who received physical therapist services from one of the Low Back Outcome team members at one of the academic medical center’s cost centers were included in the data analysis.

Downstream costs were evaluated via billing record charges before contractual adjustments. Direct physical therapist charges were evaluated via the billing record for total charges before contractual adjustments corresponding to an episode of care for physical therapist services provided to individuals with diagnoses related to LBP. Physical therapist service utilization was evaluated via the billing record for the number of physical therapist sessions.

All data were collected from physical therapists on the Low Back Outcome team who provided physical therapist services to patients at 11 outpatient clinics within the same academic health system. The pre-implementation group consisted of 21 physical therapists and the post-implementation group consisted of 16 physical therapists (14 were part of the pre-implementation group). The discrepancy in the number of physical therapists in the pre- and post-implementation groups was due to changes in physical therapists working within the health system or no longer working in the outpatient setting with patients diagnosed with LBP.

Data validation, categorization, and analysis

All samples were randomly selected from each group (pre- and post-implementation) by the medical center’s Information Warehouse. Demographic and cost-related data were collected from patient electronic billing records. Sample size estimation was calculated at 135 data points (patients) for each group using the statistical sample size estimator G*Power for Mac version 3.1 (Buchner, Erdfelder, and Faul, 1996) using a medium effect size, alpha level of 0.05, and a power level of 0.80. All categorical demographic data were analyzed using chi-squared tests of independence. Shapiro–Wilk tests for normality were performed prior to analyses on direct PT and medical costs. All cost category data were found to be non-normally distributed, however t-tests were used as all other assumptions for parametric tests were met (i.e. homogeneity of variance, independent groups, large data set, and continuous data). Mean costs for direct physical therapist services and services that occurred prior, concurrent, and downstream to physical therapist management of patients with LBP were calculated using two-tailed t-tests (Table 3). Further statistical analyses were performed to ascertain whether the statistically significant lower downstream costs in the post-implementation group were offset by the higher prior costs in the post-implementation group. Total costs for prior, concurrent, downstream, and direct physical therapist services were analyzed using two-tailed t-tests (Table 4).

Table 1. Patient demographics.

Characteristic	Pre-implementation: mean (%) N = 135	Post-implementation: mean (%) N = 135	Sig. (2-tailed)
Gender			0.17
Male	56 (41.5)	45 (33.3)
Female	79 (58.5)	90 (66.7)
Age (years)*			0.04
Average (std. dev)	54.54 (13.26)	57.55 (11.13)
Range	25–80	22–79
Race			0.6
White	83 (61.5)	72 (53.3)
African American/Black	46 (34.1)	56 (41.5)
Asian	1 (0.7)	1 (0.7)
Other/more than one race	5 (3.7)	6 (4.4)
Ethnicity			0.55
Not Hispanic or Latino	131 (97)	132 (97.8)
Latino/Hispanic Other	3 (2.2)	3 (2.2)
Refused to answer	1 (0.7)

*Analyzed with t-test, alpha level = 0.05; Sig. = Significance; Bold = statistical significance; std. dev = standard deviation; N = sample size

Table 2. Payor class, ICD-10 codes.

	Pre-implementation (%)	Post-implementation (%)	Sig. (2-tailed)
Payor Class	N = 135	N = 134	0.91
Managed care	45 (33.3)	41 (30.4)
Medicare	41 (30.4)	45 (33.3)
Medicaid	47 (34.8)	45 (33.3)
Other government	2 (1.5)	3 (2.2)
Multiple Diagnoses	N = 135	N = 135	0.81
No	61 (45.2)	63 (46.7)
Yes	74 (54.8)	72 (53.3)
Primary Diagnosis code	N = 135	N = 135	0.9
Radiculopathies	50 (37)	37 (27.4)
Low back pain	47 (34.8)	56 (41.5)
Lumbago	26 (19.3)	28 (20.7)
Other diagnoses related to lumbar spine conditions	12 (8.9)	14 (10.4)

Sig. = Significance; N = sample size.

All de-identified data were coded and categorized as having occurred prior, concurrent, or downstream to physical therapist services by date of service. Cost items were coded and categorized as direct physical therapist services or as downstream cost categories (i.e. imaging, pharmacy, emergency, surgery, other rehabilitation, and “other”). Additionally, primary diagnoses of LBP were coded and categorized into radiculopathies, low back pain, lumbago, and other diagnoses related to lumbar spine conditions. Once coded, all data were analyzed with IBM SPSS software using an a-priori alpha level of 0.05.

Results

Demographics

The overall profile of a patient with LBP in both groups was a white, non-Latino female between 55 and 58 years of age. Those in the post-implementation group were older than those in the pre-implementation group [mean (SD): 57.55 (11.13) vs. 54.54 (13.26); p = .04] (Table 1). While the difference in age groups was found to have statistical significance, a three-year difference is not considered clinically relevant or associated with age-related co-founding factors. There were no significant associations between the two groups for demographics of gender, race, ethnicity, payor class, multiple diagnoses, and primary diagnosis code (Tables 1 and 2).

Table 3. Mean costs ($ USD) pre and post implementation N = 135.

Cost category	Pre-implementation mean (std. dev)	Post- implementation mean (std. dev)	t	Sig. (2-tailed)	95% confidence interval		Cohen’s D effect size
					Lower	Upper
Prior Imaging	2307 (3428)	3043 (8819)	−0.9	0.37	−2340	867	0.13
Concurrent Imaging	2443 (4801)	2221 (4902)	0.38	0.71	−941	1385	0.28
Downstream Imaging	1938 (5260)	969 (1690)	2.04	0.04	32	1905	0.49
Prior “other”	3964 (16565)	6077 (25250)	0.81	0.42	−7230	3004	0.14
Concurrent “other”	4480 (11532)	3098 (6435)	1.22	0.23	−855	3621	0.39
Downstream “other”	5773 (21195)	1500 (3230)	2.31	0.02	627	7920	0.52
Prior Pharmacy	806 (4763)	1154 (6663)	0.49	0.62	−1739	1043	0.18
Concurrent Pharmacy	851 (3113)	1030 (6124)	0.3	0.76	−1343	985	0.2
Downstream Pharmacy	855 (3268)	211 (877)	2.21	0.03	70	1217	0.51
Prior Emergency	419.70 (1040)	739.16 (3268)	1.08	0.28	−901	262	0.11
Concurrent Emergency	666 (4217)	818 (5293)	−0.26	0.8	−1298	995	0.21
Downstream Emergency	288 (953)	203 (752)	0.81	0.42	−121	290	0.34
Prior Surgery	3642 (17279)	5248 (29653)	0.54	0.59	−7422	4209	0.17
Concurrent Surgery	3776 (14574)	1901 (8502)	1.29	0.2	−983	4736	0.4
Downstream Surgery	6246 (24834)	1447 (6570)	2.17	0.03	446	9152	0.5
Prior Other Rehabilitation	351 (2576)	522 (4078)	0.41	0.68	−988	647	0.19
Concurrent Other Rehabilitation	1080 (7615)	127 (432)	1.45	0.15	−339	2246	0.42
Downstream Other Rehabilitation	403 (2372)	316 (3095)	0.26	0.8	−574	748	0.27

Bold = statistical significance; Sig. = Significance; Std. dev = standard deviation; t = t-value

Downstream and direct PT costs

All downstream costs were lower post-implementation. Statistically significant reductions in cost post-implementation were found for downstream imaging: p = .04, 95% CI [32, 1,905], effect size = 0.49; downstream “other”: p = .02, 95% CI [627, 7,920], effect size = 0.52; downstream pharmacy p = .03, 95% CI [70, 1,217], effect size = 0.51; downstream surgical costs: p = .03, 95% CI [446, 9,152], effect size = 0.5 (Table 3).

All prior costs incurred before the initiation of physical therapist services were higher in the post-implementation group (Table 4). The difference in the mean prior costs was not statistically significant p = .43, 95% CI [−18,416, 7,815], effect size = 0.14; this was also true for concurrent costs p = .25, 95% CI [−2,873, 11,082], effect size = 0.38 between the pre- and post-implementation groups (Table 4). Post-implementation downstream costs (mean: $4,636 USD, SD: $11,592) were lower compared with the pre-implementation downstream costs (mean: $15,502 USD, SD: $53,251), and the difference was statistically significant p = .02, 95% CI [1,619, 20,115], effect size = 0.52 (Table 4).

Direct physical therapist service costs were also lower in the post-implementation group (mean: $2,863USD, SD: $1,968) compared with the pre-implementation group (mean: $3,459 USD, SD: $2,838) and the difference was statistically significant p = .05, 95% CI [11, 1,182], effect size = 0.24. Additionally, total combined costs (prior + concurrent + downstream + direct physical therapist service) were lower in the post-implementation group (mean: $ 33,477 USD, SD: $72,621) compared with the pre-implementation group (mean: $43,744 USD, SD: $75,216); however, the costs were not statistically significant p = .26, 95% CI [−7,449, 27,984], effect size = 0.39 (Table 4).

Physical therapist service utilization

The mean number of physical therapist visits was 7.98 (SD: 5.85) in the pre-implementation group and 7.74 (SD: 5.22) for the post-implementation group (Figure 1). The mean difference in physical therapist visits between the groups was a fraction of a visit (0.23) and was not statistically significant (p = .73).

Figure 1. Physical therapist service utilization, mean visits.

Discussion

The purpose of this study was to examine the financial impact of implementing an LBP CPG in outpatient physical therapist practice settings on downstream health care costs and direct physical therapist service costs. It should be noted that implementation is an ongoing process that does not have a specific end point and that the implementation activities in this study represent only the activities that were undertaken between the two pre-selected time periods labeled pre-implementation and post-implementation. While the focus of this study was on direct and downstream costs and not intended to focus on implementation, we acknowledge the great importance of a well-planned implementation strategy. Direct PT and downstream costs were reduced after the education program was implemented by the health system in this study and should be considered in future activities that aim to reduce health care costs.

The study findings of lower costs for direct physical therapist services and downstream medical costs support the notion that CPG implementation/utilization and best available evidence may reduce costs of health care for patients with LBP. The lower downstream medical costs (i.e. imaging, pharmacy, surgery, and “other”) found in our post-implementation group supports findings from previous studies that have examined cost-effectiveness of CPG adherent care for the management of LBP (Castillo et al., 2020). It is worth noting that the pre-implementation downstream costs had a larger standard deviation (SD: $53,251) in proportion to the mean (mean: $15,502 USD) than the post-implementation downstream costs (mean: $4,636 USD, SD: $11,592). The large standard deviation may be an indication of variation in charges from the health system and warrants further investigation.

Despite finding a statistically significant difference between pre- and post-implementation costs for direct physical therapist services, a wide confidence interval was noted: 95% CI [11, 1,182]. This is an indication of low precision and warrants further examination to understand the variability of the results, which could be the result of the study sample size, how the interventions were billed, the actual interventions provided or other reasons. Additionally, effect sizes ranged from small to medium, however all statistically significant findings had a medium effect size (i.e. downstream: imaging, other, pharmacy, surgery, and combined costs) except direct physical therapist costs, which had a small effect size of 0.24 (Tables 3 and 4).

Table 4. Mean combined costs ($ USD).

Cost category	Pre-implementation mean (Std. dev.)	Post-implementation mean (Std. dev.)	t	Sig. (2-tailed)	95% confidence interval		Cohen’s D effect size
					Lower	Upper
Total Prior Costs	N = 121 11484 (38753)	N = 119 16784 (66999)	−0.8	0.43	−18416	7815	0.14
Total Concurrent Costs	N = 109 13298 (34195)	N = 97 9194 (22946)	1.16	0.25	−2873	11082	0.38
Total Downstream Costs	N = 71 15502 (53251)	N = 93 4636 (11952)	2.31	0.02	1619	20115	0.52
Total Direct Physical Therapist Service Costs	N = 135 3459 (2838)	N = 135 2863 (1968)	2.01	0.05	11	1182	0.24
Total Combined Cost	43744 (75216)	33477 (72621)	1.14	0.26	−7449	27984	0.39

Bold = statistical significance; Sig. = Significance; Std. dev = standard deviation; t = t-value

A recent study that examined the association between non-adherence to CPGs and medical spending for back and shoulder injuries found non-adherence was associated with a $16,000 USD increase in spending for physical therapist services and $114,000 USD for surgery (Castillo et al., 2020). Kim et al. (2019) concluded in their retrospective study that guideline adherent care for the management of LBP resulted in lower health care costs. Kim et al. (2019) defined guideline adherence to include non-surgical management of patients newly diagnosed with LBP who did not obtain imaging within 30 days of diagnosis, and who did not obtain imaging without or before a trial of physical therapist services. Kim et al. (2019) concluded that non-adherent care was common in their study sample and the addition of imaging within 30 days of LBP diagnosis and/or obtained before the initiation of physical therapist services resulted in statistically higher costs. While the CPG examined in this study is directed at the physical therapist management of LBP and does not dictate a pathway for imaging, the role of physical therapists in ordering imaging for the management of musculoskeletal conditions is an emerging area of practice due to expansions in direct access and its impact on lowering utilization and costs (Boyles, Gorman, Pinto, and Ross, 2011; Garrity et al., 2020; Keil, Baranyi, Mehta, and Maurer, 2019) and should be considered in future studies on CPGs for LBP.

The results of this study add to the growing body of evidence that highlights the significant cost burden of managing LBP. Previous studies have examined the relationship between PT utilization, guideline adherence, and timing of services and have found reduced PT utilization when services adhere to guideline recommendations and are initiated early, thus reducing the direct costs (Childs et al., 2015; Fritz, Childs, Wainner, and Flynn, 2012; Hanney, Masaracchio, Liu, and Kolber, 2016; Hoeijenbos et al., 2005; Liu et al., 2018). It is interesting to note in this study that although the utilization of physical therapist services was not changed by the implementation of the LBP CPG, direct costs were lowered after implementation. This could be the result of eliminating ineffective interventions, such as passive modalities and interventions that do not promote physical activity and return to daily activities as is recommended in the LBP CPG (Delitto et al., 2012), thus reducing services billed, while the overall number of PT visits remained the same. Further investigation is needed to examine whether reducing the number of physical therapist visits while maintaining patient outcomes would further lower direct costs, as well as lower non-direct costs incurred by the patient in the form of travel, time away from work, and child/dependent care costs.

The management strategies utilized by the PTs in this study were based on the best available evidence at the time of data collection, in addition to their clinical expertise and the individual preferences of their patients. Since data was collected, the LBP CPG has been revised and includes updates to the original recommendations and new recommendations based on emerging evidence. Key updates in the 2021 version of the CPG include targeted interventions for subgroups of patients classified as having acute or chronic LBP (George et al., 2021). It is possible to speculate that if this study had been conducted using the updated guidance in the CPG, that costs for physical therapist services may have been lower due to more targeted and effective management of subgroups of patients with LBP, thus potentially reducing billed services and physical therapist service utilization. However, the downstream costs are less likely to be impacted as they are an indication of the overall management of patients with LBP and referral pathways outside of physical therapist practice within a health system. Evidence-informed practice is continually evolving, and the results of this study demonstrate the financial impact it can have on physical therapist practice, as well as the need to examine overall multidisciplinary management, including associated costs within a health system.

Limitations

There are several limitations to consider regarding this study. First, all analyses were performed on care that was provided during a six-month time frame, either prior to the implementation of the LBP CPG or 6 months into the implementation. This time frame was selected to represent a snapshot in time but does not fully capture all costs for each patient in the study sample.

The primary aim of this study was to examine downstream costs associated with the implementation of an LBP CPG in outpatient physical therapist practice. The economic impact of LBP can be examined from multiple perspectives, including those of patients, payors, the health system, and society. Non-direct medical costs and intangible costs from pain and suffering were not analyzed (Dunet, 2012). To capture all costs associated with LBP was outside the scope of this study. Only t-tests were performed, and we did not correct for multiple comparisons as the primary purpose of this study was to identify primary potential variables that impact downstream costs and to focus on more in-depth examination and identification of possible themes for future hypothesis driven research, quality improvement projects, and to conceptualize opportunities to improve value in physical therapist practice.

Another limitation of the study that arose during data collection was the inability to capture adherence to the LBP CPG through analysis of CPT codes. While a common method, analyzing CPT codes does not fully capture elements of the physical therapist visit or provide justification as to why certain procedures were selected over others. It also does not capture additional factors that contribute to improved patient outcomes such as the therapeutic alliance, which has previously been demonstrated to positively impact patient outcomes for individuals with chronic LBP (Ferreira et al., 2013). The data pulled from the health records did not provide sufficient detail to utilize this method. The assumption was made that CPG adherence was higher after implementation activities were undertaken and the physical therapists had greater awareness of the contents in the CPG. However, this assumption has the potential to overestimate adherence post-implementation and underestimate knowledge of evidence-informed management strategies pre-implementation.

Conclusion

Based on the results of this study, the hypothesis stating a change in downstream costs post CPG implementation is accepted. Evidence-informed practice is a vital part of physical therapist care delivery. Gaps continue to exist between research and practice. CPGs have the potential to bridge those gaps and play an important role in controlling health care spending. This is accomplished by providing evidence on effective and efficient ways to manage health conditions. As new payment models are developed in health care, clinicians, administrators, and policy makers will be prompted to find new ways to add value by decreasing costs. The results of this study suggest that implementation of an LBP CPG in outpatient physical therapist practice can have a positive impact on lowering direct physical therapist service and downstream medical costs. This was found to be especially true for interventions that are known to be ineffective for LBP, such as imaging and pharmacy, as was found in the lower post-implementation costs in this study.

The lower downstream costs in this study also highlight the importance of viewing health conditions from an interprofessional context and the impact of one service on additional service utilization. Examination of downstream costs has the potential to improve understanding of interprofessional care models as well as expose new ways to reduce the costs of healthcare.

Acknowledgments

The authors would like to acknowledge Paul Tadak, Lucas VanEtten, and the members of the medical center “Low Back Pain Team.” The project described was supported by Award Number UL1TR002733 from the National Center for Advancing Translational Sciences.

Disclosure statement

The authors report that there are no relevant competing interests to declare.

Additional information

Funding

This work was supported by the National Center for Advancing Translational Sciences [UL1TR002733].