Long term follow-up of cervical intervertebral disc herniation inpatients treated with integrated complementary and alternative medicine: a prospective case series observational study

Neck pain is a common compliant whose point prevalence is estimated at 10–18 %, with lifetime prevalence reaching 30–50 %. Prevalence of neck pain in populations aged 40 or older is approximately 20 % [1, 2]. Neck pain is also related with restricted neck movement [3], and frequently accompanied by headache, dizziness, visual impairment, tinnitus, and autonomic nervous system dysfunction [4, 5]. Frequent concurrent symptoms include upper extremity pain and neurological disorders [6], and neck pain symptoms also persist in many cases leading to work loss due to discomfort [7]. Neck-related disability is generally more serious in patients with radiating pain than pain limited to the neck area [8, 9], and the main characteristic of cervical intervertebral disc herniation (IDH) is arm pain in the region innervated at the herniated disc level and/or compressed nerve root [10, 11].

The range of available treatments for cervical IDH is vast, spanning conservative treatments to various surgical modalities. Conservative treatments include NSAIDs, oral steroids, steroid injections, patient education, rest, Thomas collars, and physical therapy [12‐14]. Surgical treatment may be considered when conservative treatment fails. Neuropathy from spinal cord compression is an absolute indication for surgery. Other indications include nerve root compression signs and related motor and sensory loss. Relative indications may involve decreased quality of life due to prolonged chronic pain [15]. While surgical treatment may benefit some patients suffering severe neurological symptoms, most studies on neuropathic pain of the spine state that the long term effects are not significant [16‐20]. Although studies on the effect of conservative treatment in cervical IDH patients have occasionally been reported, whether it is effective is yet a matter of controversy, and there is a paucity of studies on the effect of complementary and alternative medicine (CAM) treatment.

According to Benefits by Frequency of Disease data from the 2013 Korean National Health Insurance Statistical Yearbook [21], 5585 patients received treatment for cervical disc disorders for 99,582 days in outpatient care, of which 100,205 days were covered by the National Health Insurance, and medical treatment expenses eligible for reimbursement surmounted to 5,370,217 Korean Won, with 4,004,731 Korean Won reimbursed. Cervical disc disorders was the 12th most frequent reason for admission to Korean medicine hospitals, showing that it is not uncommon to receive inpatient care for cervical IDH.

Such CAM treatments as acupuncture, pharmacopuncture, herbal medicine, and manual therapy are well-sought in Korea to the aim of securing a less invasive, non-surgical method of treatment. Jaseng Hospital of Korean medicine, a Korean medicine hospital accredited by the Korean Ministry of Health and Welfare to specialize in spine disorders, treats over 900,000 spinal disease outpatient cases per year. This hospital manages patients with an integrative system utilizing conventional and Korean medicine, where conventional doctors and Korean medicine doctors (KMDs) cooperate for optimal treatment results. Conventional doctors participate in diagnosis using imaging technology such as X-rays and MRIs, and in treatment by caring for a small percentage of patients potentially in need of more intensive care. KMDs supervise and manage the main treatment of all patients, and decide whether the patient requires additional diagnosis and treatment from a conventional doctor. Cervical IDH patients suffering neck pain or radiating pain unable to receive outpatient treatment are thus provided with concentrated non-surgical integrative treatment during admission.

Despite the widespread use of inpatient treatment for cervical IDH encompassing a number of treatment modalities, studies on its treatment effect in patients admitted for cervical IDH are scarce. An integrative inpatient treatment approach with focus on CAM may not be widely available to patients, and the objective of this study is to introduce and assess the feasibility and long term effect of this integrative treatment model in inpatients with cervical IDH using a practical study design.

During the study period 784 patients with neck disorders were admitted, and of these, 234 patients were diagnosed with cervical IDH with no other major musculoskeletal complaints. Of the 234 cervical IDH patients, 175 patients had no missing values in NRS and NDI at admission and at 2 weeks post-admission or at discharge (short term follow-up). Ten patients were re-admissions and after inclusion of initial admission data if initial admission was during the study period, 165 patients remained. Long term follow-up assessments were conducted in 117 patients. In the non-long term follow-up group (n = 48), 23 patients did not answer the phone, 10 refused to participate in the long term follow-up, and 15 had since changed number or had incoming calls barred (Fig. 1). Baseline characteristics by long term follow-up group and non-long term follow-up group are listed in Table 1. Though there were no other marked differences between the 2 groups, 29 patients in the long term follow-up group had been recommended surgery (24.8 %), while only 1 patient in the non-long term follow-up group (0.02 %) had been recommended.

Average length of hospital stay was 20.8 ± 11.2 days. The majority of participants received inpatient treatment focused on Korean medicine and CAM. Herbal medicine was taken in accordance with the treatment protocol in decoction form by 81.8 % of patients and in pill form in 86.1 %, and the other patients were prescribed other herbal medicines at the KMD’s discretion. In use of conventional treatments not specified in the CAM treatment protocol, 18.2 % patients took analgesic medications or intramuscular injections an average 2.7 ± 2.3 times, and 4.8 % patients were administered 1.6 ± 0.5 epidural injections during hospital stay (Table 2). We did not implement restrictions in pharmacological treatment for study purposes, and allowed conventional medicine physicians full freedom to assess and prescribe conventional medicine as the physician deemed necessary for the patient. NSAIDs, antidepressants, and muscle relaxants were the main medicines used, and opioids were administered in the short-term in only 2 patients.

NRS of neck pain, NRS of radiating pain, and NDI all decreased significantly at discharge and at long term follow-up compared to baseline (admission) (Table 3). The major site of pain of neck and radiating arm pain showed a decrease larger than MCID (NRS decrease of 2.5 or larger in neck pain or radiating pain), and NDI scores also improved over the MCID score of 7.5 [34, 35]. Difference in NRS at discharge in the long term follow-up group (n = 117) was 2.71 (95 % CI, 2.33, 3.09) for neck pain, 2.33 (95 % CI, 1.9, 2.77) for arm pain, and that of NDI, 14.6 (95 % CI, 11.89, 17.32). Difference in NRS at long term follow-up for neck pain and arm pain from baseline was 3.15 (95 % CI, 2.67, 3.64) and 2.64 (95 % CI, 1.99, 3.29), respectively. Difference in NRS at discharge in the non-long term follow-up group (n = 48) was 2.83 (95 % CI, 2.22, 3.45) for neck pain, 2.48 for arm pain (95 % CI, 1.84, 3.12), and that of NDI was 14.86 (95 % CI, 10.41, 19.3). The between-group difference in effect between admission and discharge in the long term follow-up and non-long term follow-up patients was not significant (NRS of neck pain : p-value = 0.741; NRS of radiating arm pain: p-value = 0.646; Neck disability index: p-value = 0.775).

The average period from admission to long term follow-up was 625.36 ± 196.7 days. All 165 patients answered the PGIC at discharge, and of these patients 84.2 % replied that their state was “satisfactory” or higher. A total 117 patients replied to PGIC at long term follow-up, and 79.5 % rated their current state to be “satisfactory” or higher. PGIC was reported to be very satisfactory in 48 patients (41.0 %), satisfactory in 45 (38.5 %), slightly satisfactory in 18 (15.4 %), and dissatisfactory in 6 (5.1 %). Nine patients had undergone surgery (7.6 %), while 21 patients replied that they were currently receiving treatment. Of patients currently under treatment, 10 patients (8.5 %) continued to receive CAM, 12 patients (10.3 %) had selected conventional treatment, and 1 patient was receiving both (Table 4).

Sex, age, and unilateral radiating pain satisfied p < 0.10 in univariate analysis of baseline characteristics. Satisfaction rate increased with older age in multivariate analysis. Patients with unilateral radiating arm pain tended to be more satisfied with treatment that those without radiating pain. Also, patients receiving CAM treatment showed higher satisfaction rates than patients receiving no treatment (Table 5).

Liver function was measured in all patients at admission, and nine patients with liver enzyme abnormalities at admission received follow-up blood tests at discharge. Liver enzyme levels returned to normal in 6 patients at discharge, while 2 retained liver enzyme abnormalities, and 1 patient sustained liver injury and on further assessment was diagnosed with active hepatitis showing Hbs antigen positive and Hbs antibody negative. There were no cases of systemic immunological reactions to bee venom pharmacopuncture requiring additional treatment and no other adverse events were reported.

These results show that inpatient treatment primarily focused on CAM maintains long term effects of pain relief and functional improvement in cervical IDH patients with neck pain or radiating arm pain. NRS and NDI scores at discharge and at long term follow-up all displayed significant decrease. Also, as statistical significance and clinical significance may differ, we checked for MCID and confirmed that both NRS and NDI scores improved over MCID. MCID has been reported at 2.5 in VAS for neck pain and radiating arm pain, and 7.5 in NDI scores [34, 35]. Average improvement in pain and functionality scales all exceeded MCID, and these results are likely to be reflected in patient satisfaction rate. Out of 165 patients, 128 patients (84.2 %) rated their current state as “satisfactory” or higher at discharge. At long term follow-up, 9 (7.6 %) out of 117 patients were confirmed to have received neck surgery, and most patients showed continued decrease in NRS and NDI. In addition, 96 patients (82.1 %) currently did not receive treatment for neck pain symptoms, and 93 patients (79.5 %) replied their state was “satisfactory” or higher. As comparison of between-group difference in the long term follow-up and non-long term follow-up patients was not designed a priori, this data may be regarded to be a post hoc data analysis. The between-group difference in effect between admission and discharge in the long term follow-up and non-long term follow-up patients was not significant, and in MCID, which could be considered a more clinical measure, the 2 groups produced comparable results.

Despite the fact that all patients underwent intensive Korean medicine treatment for the duration of hospital stay, no adverse events related to treatment were reported, demonstrating the safety of integrative medicine with focus on CAM. The authors had previously conducted a retrospective study to assess safety of herbal medicine and combined intake of herbal and conventional medicine in liver function test results of 6894 inpatients hospitalized at Korean medicine hospitals, and test results of the cervical disc herniation patients included in the present study were also described [36].

A major strength of this study is that it depicts clinical practice and the results reflect treatment as it is actually practiced in Korea in conventional and Korean medicine integrative treatment settings focused on CAM. Protocol treatment was standardized and comprised of interventions whose efficacy has been confirmed in pilot studies and frequently used in clinical practice, but the protocol also allowed for individual tailoring according to patient characteristics and symptoms as seen necessary by KMDs, and the percentage and frequency of these deviations were recorded. The satisfaction rate assessed at discharge not only reflects patient attitude toward treatment effect, but also increased medical costs entailed by inclusion of various treatments. Taking into account that the participants of this study were not patients recruited through advertisements, but patients visiting a Korean medicine hospital from personal choice receiving no economic compensation for study participation, the fact that most patients’ satisfaction rate was high is particularly noteworthy. The results of this study contribute to an evidence base for superior efficacy of compositive treatment over individual treatment in patients diagnosed with cervical IDH, and verify feasibility of clinical implementation with consideration for increased compositive treatment costs.

Another limitation is that we failed to conduct more comprehensive medical evaluations. For example, although participants were diagnosed as disc herniation to be the main pathology based on MRI readings and neurological symptoms by KMDs, additional imaging information such as pathological disc level and severity of herniation were not collected. Also, data on subsequent recurrences, duration of all episodes and whether some were absolutely cured were not included in long term follow-up assessments, limiting multidimensional evaluation. In addition, while these cervical IDH patients required admission for severe neck and arm pain and consequent functional disability, the fact that this was the first attack of neck pain for many may have been cause for more favorable outcome.

However, the influence of long term follow-up compliance may not be confined to availability but potentially be associated with long term treatment effectiveness. As difference in characteristics of long term follow-up and non-long term follow-up patients may be reflected in short-term outcomes assessed at discharge and types and amount of additional conventional treatment, the fact that this study did not consider for these potential effects through additional analyses is a further limitation of this study.

Controversy still surrounds the efficacy of treatments for cervical IDH. While epidural steroid injections are the commonest modality of conservative treatment used in the United States [37] various systematic reviews show that effects are highly variable and not conclusive [38‐44]. Two approaches are widely used in epidural injections: interlaminar and transforaminal approaches. The transforaminal approach has been criticized for safety risks [45‐50], and though safer than the transforaminal approach, the interlaminar approach also holds potential risks [51‐56]. Reports on the efficacy of conventional medicine for neuropathic pain show conflicting results [57‐61], and study results on physical therapy are also inconsistent [62‐64].

Gebremariam et al. [65] evaluated the efficacy of various cervical IDH treatments in a recent review, and concluded that though the single published study on conservative treatment versus surgery showed that surgery led to better results than conservative treatment, lacking intergroup analysis, there is no evidence supporting that one treatment is more superior. Despite recommendations for initial conservative treatment and management, some patients may select surgery for cervical IDH to the main aim of alleviating radiating pain in neuropathy and preventing progression of neurological damage in myelopathy [66]. Although the evidence base of conventional conservative and surgical treatments for cervical IDH weighing the benefits and harms is somewhat insufficient, the area has been extensively studied, while there is a distinct paucity of correlative studies on CAM.

Manchikanti et al. [67] stated in a 2 year follow-up study comparing epidural injection treatment with lidocaine and a mix of lidocaine and steroids for cervical IDH that NRS in the lidocaine group was 7.9 ± 1.0 at baseline, and 3.8 ± 1.6 at the 2 year follow-up, while NRS in the lidocaine and steroid group was 7.9 ± 0.9 at baseline, and 3.8 ± 1.7 at the 2 year follow-up. NDI in the lidocaine group was 29.6 ± 5.3 at baseline, and 13.7 ± 5.7 at the 2 year follow-up, and NDI in the lidocaine and steroid group was 29.2 ± 6.1 at baseline, and 14.3 ± 6.9 at the 2 year follow-up. When compared to our study, though improvement in NRS is slightly bigger in the study by Manchikanti et al., that of NDI is similar. The baseline NRS was higher at 7.9 in this previous study, and they did not differentiate between neck pain and radiating pain in NRS assessment.

The 1 year follow-up results comparing conservative treatment and plasma disc decompression (PDD) for contained cervical IDH show that VAS scores decreased 65.73, while NDI decreased 16.7 in the PDD group (n = 61), and that VAS scores decreased 36.45, and NDI decreased 12.40 in the conservative treatment group (n = 57) [68]. However, the study subject was limited to contained cervical IDH, the outcome measure for pain was VAS preventing direct comparison, and the follow-up period was shorter than our study.

The model of integrative treatment used at a Korean medicine hospital may be highly disparate from CAM treatment models used in Western countries. Although CAM treatment is gaining widespread popularity in the West, CAM is usually limited to “complementary” rather than “alternative” medicine, and is generally practiced by conventional practitioners as an adjunctive to conventional treatment after education on acupuncture/naturopathy/etc. or through referral to CAM specialists, of whom some do not hold individual practice rights. On the other hand, Korea adopts a dual medical system where KMDs hold practice rights equal to conventional practitioners, and she does not employ a primarily family practice-based medical system, allowing patients the freedom of primary treatment selection of conventional treatment or Korean medicine treatment. The participants of this study were patients visiting and admitted to a Korean medicine hospital for Korean medicine treatment of cervical IDH, and the integrative treatment model implemented at this Korean medicine hospital does not use CAM as a supplementary measure. Therefore, treatment comprised of CAM treatment such as acupuncture, herbal medicine, Chuna manipulation, and bee-venom pharmacopuncture in most patients, and conventional treatment was administered by conventional doctors through referral in a select few. A total 18.2 % of patients received analgesic medications prescriptions 2.7 times over an average admission period of 20.8 days, which is equivalent to 1–2 days worth’s prescription (calculated as 2 times/day), and epidural injections were administered to only 4.8 %, which is low considering that these patients required admission. It can be surmised that the main objective of admission in conservative treatment for most cervical IDH patients is alleviation of pain. The fact that many inpatients displayed significant pain and functional recovery in this study holds relevance for patients considering selecting a Korean medicine hospital for conservative treatment over surgery. Also, patients were confirmed to have maintained their improved state at long term follow-up, and only 9 received surgery out of the 117 patients assessed in the long term.

Patients were divided into 2 groups by satisfaction rate as evaluated at long term follow-up with PGIC, and multivariable logistic regression analysis was conducted on baseline characteristics to assess predictive factors for satisfaction and dissatisfaction. Older age was associated with higher satisfaction rate, and unilateral radiating pain was shown to be related with higher satisfaction rates than no radiating pain. In addition, patients receiving CAM treatment were associated with higher satisfaction rates compared to those not receiving treatment. This could be partly explained by the fact that more older patients may have higher levels of pain and be in more advanced stages of degeneration, resulting in more favorable and satisfactory treatment outcomes. Similarly, patients with unilateral radiating pain suffer neurological symptoms likely to be more severe than those with no radiating pain. In addition, patients continuing to receive CAM treatment may be more favorably predisposed toward CAM, resulting in higher satisfaction rates.

While numerous prospective long term studies have been conducted on injection treatment or surgical procedures, those on CAM treatment and inpatient treatment are few. The results of this study are comparable to the prospective long term results of injection treatment. Few studies have been conducted on admission treatment for patients with a main complaint of cervical IDH, which may be related with the difference in general healthcare systems.

In conclusion, although the observational nature of this study limits us from drawing more decisive conclusions lacking a control, 3 weeks’ integrative inpatient treatment mainly comprised of CAM applied to actual clinical settings may result in satisfactory results and pain and functional improvement maintained in the long term in neck pain or radiating arm pain patients diagnosed with cervical IDH.