Abstract
Background Neck pain (NP) is prevalent and costly. NP clinical practice guidelines are similar to those for low back pain (LBP), emphasizing non-pharmaceutical and non-interventional first-line approaches. Primary care providers (PCP) are frequently consulted by individuals with NP.
Objective Examine the association between guideline concordant incorporation of non-pharmaceutical therapies, use of imaging, pharmaceutical, and interventional services, and total episode cost for individuals with NP initially contacting a PCP.
Design Retrospective cohort study using identical methods to a previous LBP study
Setting/Patients National sample of individuals with non-surgical NP occurring in 2017-2019.
Measurements Independent variables were initial contact with a PCP, and the timing of incorporation of 5 types of non-pharmaceutical therapies. Dependent measures included use of 13 types of health care services and total episode cost.
Results 70,252 PCPs were initially contacted by 124,780 individuals with 137,274 episodes of non-surgical NP. 30.9% of PCPs and 22.1% of episodes included at least one of five non-pharmaceutical services at any time during an episode. Active care (13.7% of episodes), manual therapy (10.8%), and chiropractic manipulative therapy (9.4%) were the most common non-pharmaceutical services. 7.4% of episodes included a non-pharmaceutical service during the first 7 days with these episodes associated with a modest reduction (risk ratio 0.28 to 0.78) in the use of prescription pharmaceuticals. Younger individuals from ZIP codes with higher adjusted gross income were more likely to receive a non-pharmaceutical service in the first 7 days of an episode. When included during an episode, non-pharmaceutical services were associated with an increase in total episode cost with the smallest increase associated with chiropractic and osteopathic manipulation.
Limitations As a retrospective observational analysis of associations there are numerous potential confounders and limitations.
Conclusions Non-pharmaceutical services are infrequently provided to individuals with non-surgical NP initially contacting a PCP. For these individuals, non-pharmaceutical services, if provided, are most commonly introduced later in an episode after receiving pharmaceutical, imaging, and interventional services. For individuals with NP initially contacting a PCP there is an opportunity to increase the guideline concordant incorporation of non-pharmaceutical services early in an episode.
Introduction
Neck pain (NP) is prevalent, costly, and associated with a high number of years lived with disability.1-7 NP is a common reason for a visit to a primary care physician (PCP).8-10 After chiropractors PCPs are the second most common type of health care provider (HCP) initially contacted by individuals with NP.11
Among spinal disorders, management of low back pain (LBP) benefits from the availability of relatively homogenous high-quality clinical practice guidelines (CPGs) that describe a stepped approach in which services are sequenced into first-, second- and third-line services.12-14 In the absence of red flags of serious pathology, LBP CPGs emphasize individual self-management, non-pharmaceutical, and non-interventional services as first-line approaches.12-14 NP CPGs, while less common and more heterogenous than CPGs for LBP, generally recommend a similar approach where non-pharmaceutical and non-interventional approaches are recommended as initial management options.15-26
Spinal disorders, particularly LBP, have been identified as a source of a high proportion of non-guideline concordant “low-value” care, described as services generating costs without or with minimal beneficial impact on outcomes.27,28 The time burden associated with providing CPG concordant care, including for NP, is considerable.8,29,30 Examples of low-value care for LBP include overuse of imaging, interventional procedures, and some prescription medications such as opioids.31-34 The lower prevalence of NP and heterogeneity of NP CPGs results in less being known about the magnitude of low-value care for NP.
The type of HCP initially contacted has been used as a method to evaluate variation in service utilization and cost outcomes for LBP and NP.11,35-38 When initially contacted by an individual with NP, PCPs generally incorporate imaging and pharmaceutical services more frequently than CPG recommended non-pharmaceutical therapies.35 Several barriers to PCP referral for non-pharmaceutical therapies have been identified for spinal disorders.39-46 These barriers include coverage limitations 39,45, inconvenient access 40,45, cost 40,41,45, lack of familiarity and communication 42-44, limited time to make a referral 45, and concerns about possible adverse events.46
The aim of this retrospective, observational study was to replicate the methods used in an earlier study of LBP 47 to examine the associations between frequency and timing of incorporation of active care (AC), manual therapy (MT), chiropractic manipulative therapy (CMT), osteopathic manipulative therapy (OMT), or acupuncture (Acu) services, utilization of other healthcare services, and total episode cost for individuals with non-surgical NP initially contacting a PCP. Like the previous LBP study 47 the hypothesis was that early incorporation of one or more non-pharmaceutical services would be associated with lower rates of imaging, pharmaceutical and interventional service use, and lower total episode cost.
Methods
Study design, population, setting and data sources
This was a retrospective cohort study of individuals initially contacting a PCP for non-surgical NP. The study design was identical to a previous study of LBP conducted by the same author group.47 De-identified enrollment records and administrative claims data for individuals with NP were included in an enrollee database. HCP de-identified demographic information and professional licensure status was included in a HCP database. ZIP code level adjusted gross income (AGI) data was extracted from the Internal Revenue Service 48, population race and ethnicity data from the US Census Bureau 49, and socioeconomic status (SES) Area Deprivation Index (ADI) data, from the University of Wisconsin Neighborhood Atlas® database.50
With study data de-identified or a Limited Data Set in compliance with the Health Insurance Portability and Accountability Act (HIPAA) and customer requirements the UnitedHealth Group Office of Human Research Affairs determined that this study was exempt from Institutional Review Board review. The study was conducted and reported based on the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.51 [Supplement – STROBE Checklist]
With the available data it was not possible to differentiate whether non-pharmaceutical services were provided by a HCP accessed directly by the individual with NP after initially contacting the PCP or provided following referral from the either PCP or another HCP. The standard practice of attempting to generate causal inferences by adjusting for measurable, yet incomplete, confounders such as age, sex and co-morbidities 52,53 using potentially inadequate approaches such as propensity score matching, 54 not only does not yield causal insights, the distorted results can potentially limit translation potential. As an alternative, in this observational study actual demographic and episodic measures and associations are reported to facilitate straightforward translation. Examples of unmeasurable and potentially important confounders include: nuanced clinical complexity of NP not captured in administrative data, anticipated potential out of pocket costs and individual willingness to pay for different services, individuals self-paying for non-pharmaceutical or other services, availability of HCPs offering non-pharmaceutical services convenient to an individual’s home, workplace or daily travel routes, individual preference for specific services or types of HCP including gender or racial concordance, recommendations from family or friends, and appointment availability within a PCP’s and individual’s timing expectations for HCPs meeting these and other criteria.55
Cohort selection and unit of analysis
The cohort included individuals aged 18 years and older initially contacting a PCP for a complete episode of NP commencing and ending between 1/1/2017 and 12/31/2019. This timeframe was selected as the period before the influence of the COVID-19 epidemic on care patterns in early 2020. All individuals had continuous medical and pharmacy insurance coverage during the entire study period.
Administrative claims data were translated into the episode of care unit of analysis using the Symmetry® Episode Treatment Groups® (ETG®) and Episode Risk Groups® (ERG®) version 9.5 methodologies and definitions. ETG® and ERG® have been reported as a valid measurement for comparison of HCPs based on cost of care56, with previous studies finding a low risk of misclassification bias associated with the episode of care unit of analysis.11,35,47
Complete episodes, defined as having at least 91-day pre- and 61-day post-episode clean periods, were included in the analysis. Incomplete episodes were excluded. Also excluded were NP episodes including a surgical procedure, or episodes associated with diagnoses of malignant and non-malignant neoplasms, fractures and other spinal trauma, infection, congenital deformities and scoliosis, autoimmune disorders, osteoporosis, and advanced arthritis. As an observational analysis of associations, these straightforward exclusions were made to address a potential study limitation of individuals with more complex NP conditions influencing the timing of incorporation of first-line non-pharmaceutical and non-interventional services. By removing more complex diagnoses the analysis was able to focus on less complicated NP.
Variables
Data preprocessing, table generation, and initial analyses were performed using Python (Python Language Reference, Version 3.7.5., n.d.). A goodness of fit analysis was conducted using D’Agostino’s K-squared test. Non-normally distributed data are reported using the median and interquartile range (IQR).
The primary independent variables were initial contact with a PCP, and the timing of incorporation of AC, MT, CMT, OMT, or acupuncture (Acu) services. The following Current Procedural Terminology (CPT®) codes were used to identify non-pharmaceutical services: AC – 97110, 97112, 97530; MT – 97140; CMT – 98940 to 98942; OMT – 98925 to 98929; Acu– 97810, 97811, 97813, 97814. For NP, these are the most frequently provided non-pharmaceutical services recommended by CPGs and covered by commercial insurance.35 For this study passive therapies were excluded from the definition of non-pharmaceutical services. The timing of incorporation of AC, MT, CMT, OMT, or Acu services was based on the number of days after the initial visit with a PCP when a non-pharmaceutical service was first billed by any HCP.11,35,47
The PCP HCP category consisted of Family Practice, Internal Medicine, General Medicine, and OBGYN physician types, along with Nurse Practitioner and Physician Assistant HCPs. The study cohort was able to access all PCP HCP types directly without a referral.11,35,47
The primary dependent variable was the rate and timing of use of 13 types of health care services.11,35,47 Secondary dependent variables included the total cost of care for all reimbursed services provided by any HCP during an episode, the number of different HCPs seen during an episode, and episode duration measured in days. Total episode cost included costs associated with all services provided for an episode of NP, including those not specifically identified in the 13 categories used in the analyses. Costs for services for which an insurance claim was not submitted were not available. The episode duration was the number of days between the first and last date of service for each episode.
Risk (RR) ratios, and associated 95% confidence intervals (CI), were calculated for the timing of introduction of each non-pharmaceutical service type compared to a baseline of a specific non-pharmaceutical service not being provided. Due to odds ratios tendency to exaggerate risk in common outcomes, RR were reported as the measure more widely understood in associational analyses.57 Bivariate analyses were also performed comparing episode attributes associated with timing of introduction of each non-pharmaceutical service type. Like the RR calculation the bivariate analysis baseline was episodes that did not include a specific non-pharmaceutical service. Differences in the percent of episodes including a service was evaluated using Fisher’s Exact test (p value of .001), with measures reported using median and IQR evaluated using Mann Whitney U test (p value of .001).
Role of Funding Source
None
Results
The sample included 124,780 individuals, with a median age of 47 (Q1 37, Q3 55), and 61.0% females. These individuals were associated with 137,274 complete non-surgical NP episodes involving 70,252 unique PCPs. There were $79,712,281 in reimbursed health care expenditures with a median total cost per episode of $157 (Q1 $51, Q3 $495). The median pre-episode clean period was 634 days (Q1 423, Q3 858). The median number of days between sequential episodes was 209 (Q1 119, Q3 346). The median post-episode clean period was 432 days (Q1 264, Q3 684) [Table 1]. Individuals were from all 50 States and some U.S. territories. [Supplement – State].
77.9% of non-surgical NP episodes did not include a non-pharmaceutical service at any time during an episode. For the 22.1% of episodes that included any non-pharmaceutical service at any time during an episode, AC (13.7% of episodes), MT (10.8%) and CMT (9.4%) were most common. OMT (2.3% of episodes) and Acu (0.4%) were infrequently provided at any time during an episode. Individuals were more likely to receive skeletal muscle relaxants (34.8% of episodes), prescription NSAIDs (30.5%), radiography (25.8%), and opioids (15.7%) than non-pharmaceutical services. [Table 2]
Within the first 7 days of initially contacting a PCP 7.4% of episodes of non-surgical NP included one or more of the five non-pharmaceutical services with AC (3.7% of episodes), MT (2.8%) and CMT (2.1%) being most common. When introduced in the first 7 days after initial contact with a PCP, non-pharmaceutical services were generally associated with a reduction in exposure to prescription pharmaceuticals and an increase in exposure to radiology, MRI, and spinal injection services. OMT was associated with the largest reduction in exposure to prescription opioids (RR 0.29, 95% CI 0.25-0.25) followed by acupuncture (0.62, 0.39-.099), AC (0.63, 0.58-0.69), MT (0.65, 0.59-0.71) and CMT (0.77, 0.70-0.85). A non-pharmaceutical service introduced 8-14 days after initial contact with a PCP was associated with less significant and generally not clinically meaningful reduction in exposure to prescription NSAIDs and opioids along with an increase in exposure to spinal imaging and injections. When a non-pharmaceutical service was introduced 15+ days after initial contact with a PCP exposure to prescription pharmaceutical, spinal imaging and spinal injections were generally higher than if a non-pharmaceutical service was never provided [Table 2] [Supplement – Risk Ratio]. The RR for exposure to second- and third-line services based on timing of introduction of AC is illustrated in Figure 1, and CMT in Figure 2.
Individuals with non-surgical neck pain initially contacting a primary care provider. Risk ratio and 95% confidence interval for exposure to various health care services based on timing of introduction of active care compared to episodes without active care.
Individuals with non-surgical neck pain initially contacting a primary care provider. Risk ratio and 95% confidence interval for exposure to various health care services based on timing of introduction of chiropractic manipulative therapy (CMT) compared to episodes without CMT.
Episodes with a non-pharmaceutical service introduced within 7 days of initial contact with a PCP were associated with younger individuals, with a lower ERG® risk score, from zip codes with lower deprivation, higher AGI, and greater availability of a chiropractor (DC) or physical therapist (PT). Among individual non-pharmaceutical services, Acu was most strongly associated with lower deprivation, higher AGI, lower percent non-Hispanic white population, and greater availability of a licensed acupuncturist (LAc) [Figure 3] [Table 3].
For individuals with neck pain initially contacting a PCP, Area Deprivation Index (ADI) of the individual’s home address zip code associated with median (Q1, Q3) number of days (d) into an episode when first line services are initially introduced
Compared to episodes without a non-pharmaceutical service, total episode cost was higher when any non-pharmaceutical service was provided at any time, except for OMT provided within 7 days of initial contact with a PCP. The total episode cost increase was lowest for introduction of CMT and OMT [Figure 4]. Episode duration increased as non-pharmaceutical services were introduced later in an episode [Table 3].
For individuals with neck pain initially contacting a PCP, median (Q1, Q3) total episode cost associated with number of days (d) into an episode when first line services are initially introduced
Discussion
PCPs play an important role in the health care delivery system and are initially contacted by individuals with a variety of conditions, including NP. In the absence of serious pathology, NP CPGs generally recommend non-pharmaceutical and non-interventional approaches. Time burden, administrative complexity, cost, and other factors likely contribute to the observed low rate of use of CPG recommended non-pharmaceutical and non-interventional approaches for individuals with NP initially contacting a PCP. Strategies and tactics that make it easier for PCPs to provide CPG concordant care for NP is an important focus of translation efforts.
As a retrospective cohort study of associations, there are numerous potential limitations and confounders. The risk of selection bias was present due to the limited ability to control for individual preference and/or meaningful differences in clinical complexity of individuals seeking treatment of NP by a PCP versus other types of HCP, and individual expectations or requests for specific types of health care services. These limitations were partially addressed by limiting the cohort to only those individuals with non-surgical NP, and by excluding NP episodes associated with significant pathology.
Data errors, variability in benefit plan design, variability in enrollee cost-sharing responsibility, and missing information were potential sources of confounding or bias. These were partially addressed through limiting the cohort to only those having continuous highly uniform commercial insurance coverage and the processing of administrative claims data having included extensive quality and actuarial control measures. The identification of individual PCPs was based on data contained in a commercial insurer HCP database that may have included errors or missing information. Summarizing total episode cost has potential limitations associated with insurance coverage, nature of network participation, alternative reimbursement models, and individuals obtaining services outside of insurance coverage and reimbursement. The cohort, while including individuals from all 50 states and most US territories did not describe a U.S representative sample.
This study corroborates and expands on two earlier studies. First, an identical study of individuals with LBP initially contacting a PCP also found low rates of incorporation of AC, MT, CMT, OMT, and Acu.47 Second, a previous study found a low proportion of individuals with NP initially contacting a PCP have timely incorporation of guideline-concordant non-pharmacologic and non-interventional therapies.35 This study expands on this to explore more detailed distribution of episodes by timing of incorporation of AC, MT, CMT, OMT, and Acu. While these services are infrequently incorporated in NP episodes initially contacting a PCP, when incorporated AC, MT, CMT, OMT, and Acu tend to be introduced after 14 days into an episode. Future research should explore the degree to which patient self-selection versus PCP referral results in incorporation of AC, MT, CMT, OMT, and Acu for episodes of NP.
A previous study of spine-related disorders (SRDs), including NP, compared usual primary care, with a “primary spine care” (PSC) model.58 The PSC model was comprised of DCs, PTs and DOs embedded within a traditional primary care setting and directly accessed by individuals with a SRD. Compared to usual primary care, the PSC model was found to be associated with lower total costs in both the first and second year, with no meaningful differences in clinical outcomes. In this study, total costs were found to be higher when individuals with NP initially contacted a PCP and subsequently obtained treatment from a DC, DO, PT, or LAc. The absence of clinically meaningful differences in pharmaceutical, imaging, or interventional services indicates AC, MT, CMT, OMT, and Acu services were additive to typical primary care management. Further research comparing attributes of direct versus referral-based access to AC, MT, CMT, OMT, and Acu is warranted.
Previous studies of PCP referral patterns for LBP and chronic musculoskeletal pain found that administrative burden and the cost of non-pharmacologic therapies are perceived as barriers.41,42 This appears to be corroborated by this study’s finding that AC, MT, CMT, OMT, and Acu are infrequently incorporated by PCPs into management of NP, and if incorporated tend to be later in an episode after pharmaceutical, imaging, or interventional services. It was not possible to know whether AC, MT, CMT, OMT, and Acu were the result of PCP referral or individuals directly accessing these services after initially contacting a PCP.
This study of individuals with NP initially contacting a PCP is similar to previous LBP studies that found earlier use of non-pharmaceutical therapies may be associated with a reduction in use of low value services and prescription pharmaceuticals, including opioids. 36,59,60 The study finding that the benefits of early use of non-pharmaceutical therapies for NP are most evident if initiated within 7 days of initially contacting a PCP corroborates a similar finding for LBP 61, has potentially important translation implications, and warrants additional study given the relative absence of data for NP.
Conclusion
A PCP is commonly the initial HCP consulted by individuals with NP or LBP. For individuals without red flags of serious pathology NP and LBP CPGs emphasize favorable natural history, self-care, and non-pharmaceutical services as first-line approaches. This study reveals that like LBP, individuals with NP initially contacting a PCP commonly receive pharmaceutical, imaging, and interventional services before non-pharmaceutical services. Non-pharmaceutical services are infrequently provided early in an episode when the potential benefits are greatest. Within the time constraints of a typical PCP visit, increasing early incorporation of guideline concordant non-pharmaceutical services likely involves making it easier for a PCP to address a variety of individual preferences, local socioeconomic, and HCP availability factors. As with LBP, making it easier for PCPs to suggest and individuals to follow through on recommendations to incorporate non-pharmaceutical services may be enhanced by a plain language summary of NP CPGs available to individuals before a visit with a PCP.
Data Availability
All data produced in the present work are contained in the manuscript
Acknowledgements
Footnotes
Declarations
Ethics approval and consent to participate Because the data was de-identified or a Limited Data Set in compliance with the Health Insurance Portability and Accountability Act and customer requirements, the UnitedHealth Group Office of Human Research Affairs determined that this study was exempt from Institutional Review Board review.
Consent for publication Not applicable
Availability of data and materials The data are proprietary and are not available for public use but, under certain conditions, may be made available to editors and their approved auditors under a data-use agreement to confirm the findings of the current study.
Competing interests At the time of manuscript submission DE and MZ are UnitedHealth Group employees and UNH stockholders. No other potential conflicts of interest or competing interests exist.
Funding None
List of Abbreviations
- NP
- Neck pain
- US
- United States
- CPG
- Clinical practice guideline
- PCP
- Primary care provider
- PS
- Physician specialist
- DC
- Doctor of Chiropractic
- PT
- Physical Therapist
- HCP
- Health care provider I
- QR
- Interquartile range
- OR
- Odds ratio
- RR
- Risk ratio
- AGI
- Adjusted Gross Income
- ADI
- Area Deprivation Index
- STROBE
- Strengthening the Reporting of Observational Studies in Epidemiology
- CPT®
- Current Procedural Terminology
- ETG®
- Episode Treatment Group®
- ERG®
- Episode Risk Group®
- ACP
- American College of Physicians
- PA
- Physician Assistant
- CMT
- Chiropractic manipulative treatment
- OMT
- Osteopathic manipulative treatment
- AC
- Active care
- MT
- Manual therapy