A multidisciplinary pain management program for patients with chronic low back pain: a randomized, single-blind, controlled, feasibility study

Abstract

Background

Multidisciplinary programs are the first recommendation for non-specific chronic low-back pain, but implementing this type of program is complicated to get up and running. The primary aim of this study was to assess the feasibility and appropriateness of the PAINDOC multidisciplinary program for subjects with chronic low-back pain. The secondary objectives were to evaluate the decrease in pain intensity, pain-related disability and pain catastrophizing, as well as the improvement in quality of life with this program. Furthermore, another of the secondary objectives was to calculate the sample size for a future randomized clinical trial.

Methods

This study was conducted in a hospital pain unit using two successive recruitment waves. First, the feasibility outcomes (recruitment, completion, and drop-out rates) of a 5-month non-random prospective cohort (n = 227) were recorded. Then, the clinical outcomes (pain intensity, quality of life, disability, and pain catastrophizing) were recorded from a prospective, controlled, two-armed and single-blind feasibility study (ClinicalTrials.gov, NCT05974072). It included 41 participants that were randomly allocated to either the pharmacological treatment (n = 21) arm or PAINDOC program (n = 20) arm.

Results

The recruitment rate was 66%, with the completion rate standing at 80.7% and the drop-out rate at 19.3%. Significant differences and a medium size effect were observed between groups in terms of pain intensity (p = .017, r = .408) at the 4-month follow-up. The intragroup analysis of the PAINDOC group revealed significant lessening in pain intensity (p = < 0.001) and improvements in quality of life (p = .030).

Conclusions

This study showed that the PAINDOC multidisciplinary program is a feasible treatment for patients with non-specific chronic low-back pain. Furthermore, the exploratory results of this study suggest that it could be an effective treatment to reduce pain intensity and improve on self-reported quality of life in these patients, although a future randomized clinical trial is needed to determine its effectiveness.

Trial registration

NCT05974072 (registration date July 11, 2023; retrospectively registered; ClinicalTrials.gov).

Supplementary Information

The online version contains supplementary material available at 10.1186/s12891-025-08294-8.

Background

Chronic pain (CP) affects up to 33% of the adult population worldwide [1], with chronic low-back pain (CLBP) being the most prevalent disorder. Moreover, it is a leading cause of years lived with disability [2]. Despite increasing pharmacological and surgical medical treatments, CLBP rates continue to rise [3] and entail high economic and social costs [4].

The current recommendation is to approach CLBP from a biopsychosocial model, which is far from the traditional biomedical model focused solely on nociception [5]. This model postulates that CLBP is a dynamic interaction between social, psychological, and biological factors [4, 6] that condition certain learnings, beliefs and behaviors that contribute to the persistence of pain and disability [7, 8].

Regarding this biopsychosocial model, several studies have shown that pain catastrophizing (PC) is one of the most significant predictors of the development, progression, and perpetuation of CLBP [9, 10]. PC is a cognitive-emotional response pattern that involves negative expectations and subjective appraisal of actual or anticipated pain [11]. Accordingly, treating PC could favorably alter brain function and structure, and contribute to improve the disability status [12–14].

Therefore, there is a need for multimodal interventions that address CLBP through a biopsychosocial approach. They might include disciplines like pain neuroscience education (PNE), therapeutic exercise (TE), cognitive behavioral therapy (CBT)-based psychotherapy, therapeutic education, and mindfulness meditation. These are five therapies that, in isolation, have been shown to lessen pain and improve quality of life (QoL), among other effects [15–19]. However, multidisciplinary treatments can be complex in terms of resource availability, the coordination of healthcare professionals and patient adherence to treatment.

In this line, O’Sullivan et al. described Cognitive Functional Therapy, which is a flexible integrated behavioral approach designed to provide individualized treatment to each patient suffering from disabling CLBP [20]. This therapy combines different strategies, seems to be effective in the long term and less costly than usual care in its population [21]. However, it entails a complete individualized assessment of each patient and therefore requires considerable resources. Other previous studies looked at single interventions or combinations of a few disciplines with approaches that did not incorporate as many therapies for CLBP as our program [22, 23]. To our knowledge, there has never been a multimodal approach combining these five disciplines, specifically targeting CLBP, in a single working group that includes physicians, nurses, physiotherapists and psychologists. We propose to implement, for the first time in a Spanish hospital pain unit, a therapeutic program that simplifies the circuits for both professionals and patients.

Accordingly, the primary aim of the present study was to assess the feasibility and appropriateness of a multidisciplinary treatment program, denominated herein the Comprehensive Care Program for People with Chronic Pain (PAINDOC), as a therapeutic option for patients with CLBP. The secondary objectives were to evaluate the potential effect of the program on lessening pain intensity, CLBP-related disability, PC, and improving QoL in comparison with a pharmacological control group, as well as to estimate the sample size needed for a full randomized clinical trial. The authors hypothesized that the PAINDOC multidisciplinary program would be a feasible and potentially effective intervention for patients with CLBP.

Methods

Study design

The present study was conducted over two successive recruitment waves. First off, feasibility outcomes were recorded from a non-random, prospective cohort over five consecutive months going from November 2021 through March 2022. Then, clinical outcomes were recorded from a randomized, single-blind, controlled, parallel-group study, which was conducted from January 2023 until October 2023. The present study was registered at clinicaltrials.org (NCT05974072; registration date July 2023, retrospectively registered) and approved by the Research Ethics Committee of the Hospital Clínic de Barcelona (HCB/2023/0177). All patients included in this trial gave written informed consent to participate. This clinical trial was performed according to the Declaration of Helsinki and Good Clinical Practice Guidelines (ICH E6 R2). Patient confidentiality was ensured in accordance with current Spanish legislation (LOPD 3/2018). This manuscript complies with the CONSORT 2010 statement [24].

Participants

Patients with non-specific CLBP referred to the Pain Unit of the Hospital Clínic de Barcelona were recruited and assessed for eligibility by physicians from the Pain Unit. The inclusion criteria were that patients had to be over 18 years of age, diagnosed with primary CLBP (non-specific CLBP) of axial predominance, experiencing pain at least 50% of the time over the last 6 months [25]. They also had to have an average weekly pain intensity equal to or greater than 4 out of 10 on a verbal numerical pain rating scale (moderate to severe intensity) [26], and be agreeable to receiving active, non-pharmacological and non-surgical treatment. The exclusion criteria were having low-back pain of inflammatory origin, a history of cancer in the last 5 years, unexplained and unintentional weight loss of 10 kg or more over the preceding 12-month period, a lack of control of bowel and bladder function (cauda equina) and difficulty in performing the sessions due to severe physical disability. Having sought compensation or entered into litigation in the preceding 12-month period, having problems in understanding the content of the sessions (language barrier, severe hearing loss or severe cognitive impairment), being assessed as having a low probability of adherence to the program due to a diagnosis of severe mental illness (schizophrenia, major depression, bipolar disorder, etc.), a strong prescription opioid or parenteral medication addiction disorder, or technical-logistical problems (inability to attend treatment sessions or inability to fill out assessment questionnaires) were also criteria for exclusion.

The second-wave of recruited subjects were randomized into two groups: the Control group, which received pharmacological treatment [27, 28], and the Experimental group, which participated in the PAINDOC program.

Sample size

Since this was a feasibility study, one of the objectives was to calculate the sample needed for a randomized clinical trial. Thus, no formal sample size calculation was made. Therefore, a total of 20 participants were estimated per group (40 participants in total), which is considered adequate for feasibility studies [29].

Treatment allocation

Randomization was based on a permuted block allocation sequence using a web-based random number generator (GraphPad). Numbered, opaque sealed envelopes were used to ensure a balance of subjects in each intervention group [30]. The size of the blocks was 4, with a total of 10 blocks (40 consecutive participants). A researcher not involved in the recruitment or treatment of the patients opened the envelopes and proceeded with treatment allocation. Eventually, 41 participants were included as the last two were recruited at the same time by two different physicians.

Interventions

Control group - pharmacological control

The control group subjects were treated using a pharmacological approach that followed the latest clinical guidelines for patients with CLBP. Thus, the pharmacological options that were considered for each patient were those included in the first and second steps of the WHO Analgesic Ladder (preferably without including weak opioids, i.e., if possible, without escalating to the second step) [31]. The first step includes non-opioid analgesics (paracetamol, non-steroidal anti-inflammatory drugs and derivatives, and metamizole) and co-analgesics (antidepressants, muscle relaxants, anticonvulsants, corticosteroids, local anesthetics). The second step includes all those in the first step plus weak opioids (tramadol and codeine). Pharmacological treatment was individualized according to pain intensity, contraindications, and the onset of adverse effects [27, 28].

Experimental group - PAINDOC program

Patients randomized to the experimental group participated in the PAINDOC program. This is a multimodal and multidisciplinary treatment that encompasses five different parts delivered by different health professionals in the Pain Unit. It consists of 8 face-to-face group sessions carried out over 2 months. It includes a therapeutic education session called Empowered Relief^® that is imparted by a certified physician (anesthesiologist, rheumatologist or family physician), a pain psychology session imparted by a psychologist, an introductory session to mindfulness meditation imparted by an advanced practice nurse, and 2 sessions of PNE and 3 sessions of TE imparted by a physiotherapist (See Fig. 1. Schedule of the PAINDOC program). These sessions provide patients with self-management tools and teach them new skills to aid in coping with pain more actively. All this new information is oriented toward changing their beliefs, fears, behaviors and, thereby, reducing pain and its catastrophizing. This program had already been included in the unit’s standard care practice. Thus, the sessions did not call for an additional, specific visit for the patients in the study.

Fig. 1.

Schedule of the PAINDOC program

A complete description of each intervention is presented in an additional file [See Additional file 1].

Outcomes

Data collection

Demographic and clinical characteristics were collected at baseline prior to randomization. Three assessments were conducted. The first was at baseline. The second was at two months, which coincided with the end of the program for the experimental group. The last one was done at four months (See Fig. 2. Flow chart diagram). Evaluation appointments were scheduled by telephone.

Fig. 2.

Flow chart diagram of patients throughout the course of the study

Clinical outcomes were collected by a blinded physician from the Pain Unit, using self-administered questionnaires. They were administered in physical format, with an average response time of 30 min. Due to the very nature of the intervention, neither the therapists delivering the sessions, nor the participants could be blinded.

Demographic and clinical characteristics

Demographic and clinical data that included age, gender, body mass index, smoking habit, occupation, educational level, location and distribution of pain, and pharmacological pain treatment were collected at baseline.

Feasibility outcomes

Three feasibility outcomes were considered. One of them was the recruitment rate that was assessed by calculating the number of participants included in the program for 5 consecutive months. The number of eligible, included and excluded participants, and the reasons for exclusion were registered. The number of participants completing treatment was also assessed. Acceptable compliance is established as at least 80% of participants included in the program completing all program sessions [32]. Moreover, treatment drop-out rates were also recorded.

Secondary outcomes

The secondary outcomes consisted of the following four elements.

Clinical impact on pain intensity

The participants were asked to rate their average pain in the ongoing week using a 100 mm visual analogue pain rating scale (VAS). Here, 0 indicates “no pain” while 100 indicates “the worst pain imaginable” [33]. Pain intensity reduction was considered clinically relevant if their VAS pain score decreased by at least 20 mm [34].

Disability related to low-back pain

It was evaluated using the Spanish version of the Oswestry Disability Index (ODI). It comprises 10 items scored from 0 to 5 (higher values indicate greater disability), corresponding to 10 domains. The sum of the 10 scores is presented as a percentage (ODI score). It ranges from 0% (no disability) to 100% (maximum disability) [35].

Quality of Life (QoL): This one was assessed using the 5-dimension, 5-level, self-administered Euro Quality of Life questionnaire (EQ-5D-5 L). It consists of two pages, the EQ-5D-5 L descriptive system (page 1) and the 20 cm EQ vertical visual scale (page 2). The descriptive system comprises 5 dimensions (mobility, self-care, activities of daily living, pain/discomfort, anxiety/depression). Each dimension consists of 5 levels: no problems, mild problems, moderate problems, severe problems, and extreme problems [36].

Pain catastrophizing (PC)

The Spanish version of the 13-item self-administered Pain Catastrophizing Scale (PCS) was used for this evaluation [37]. Each item has 5 score levels ranging from (0) “not at all” to (4) “all the time”, with a total range from 0 to 52 points. The higher the score, the higher the degree of PC [38].

Statistical analysis

The Statistical Package for the Social Sciences (SPSS 28, SPSS Inc, Chicago, IL) was used for statistical analysis. The descriptive statistics, which include means and standard deviations for continuous variables and numbers and percentages for categorical variables, were used to summarize baseline data of both groups. The distribution of measures was evaluated using the Shapiro-Wilk test (p < .05). Due to the lack of normal distribution for most variables, the authors used nonparametric statistics. Continuous variables were compared using the Mann-Whitney U test while categorical variables were compared using the Chi-squared or linear Chi-squared test, depending on the number of categories. For between-group comparisons, the Mann-Whitney U test was used. And, for intra-group analysis, Friedmann’s ANOVA was computed, and a post hoc pairwise analysis was performed using the Wilcoxon signed rank test. The non-parametric effect size r was estimated for both between and within-group comparisons [39]. The r values were considered small when they were greater than 0.1, medium when > 0.3 and large when > 0.5 [40]. P values below 0.05 were considered statistically significant.

Results

Baseline demographic and clinical variables are presented in Table 1. No differences were found between the groups at baseline.

Table 1.

Baseline characteristics and clinical variables of the randomized study (second wave) participants

Variable	Total Sample (N = 41)	Control Group (n = 21)	PAINDOC Group (n = 20)	p-value
Age, (Mean, SD)	63.96 (37.4)	64.47 (35.0)	63.65 (37.1)	0.873
Gender, no. (%) women	27 (63.4)	12 (57.1)	14 (70.0)	0.585
Body mass index classification, no. (%) of participants				0.792
Underweight	1 (2.4)	0 (0)	1 (5.00)
Normal weight	10 (24.4)	7 (33.3)	3 (15.0)
Overweight	17 (41.5)	9 (42.9)	8 (40.0)
Obesity	13 (31.7)	5 (23.8)	8 (40.0)
Smoking habit, no. (%) of participants				0.507
Never smoked	15 (36.6)	8 (38.1)	7 (35.0)
Quit smoking	13 (31.7)	7 (33.3)	6 (30.0
Smoker	13 (31.7)	6 (28.6)	7 (35.0)
Occupation, no. (%) of participants				0.898
Employed	12 (29.3)	6 (28.6)	6 (30.0)
Retired	17 (41.5)	10 (47.6)	7 (35.0)
Household chores	7 (17.1)	3 (14.3)	4 (20.0)
Student	1 (2.4)	1 (4.8)	0 (0)
Unable to work	4 (9.8)	1 (4.8)	3 (15.0)
Educational level, no. (%) of participants				0.910
Read and write	1 (2.4)	1 (4.8)	0 (0)
Elementary, intermediate	14 (34.1)	5 (23.8)	9 (45.0)
Secondary, vocational	18 (43.9)	8 (38.1)	10 (50.0)
University	8 (19.5)	7 (33.3)	1 (5)
VAS pain, scored from 0 to 10 (median, IQR)	7 (5)	6.5 (5.0)	7.5 (2.75)	0.304
ODI, scored from 0 to 100 (median, IQR)	32.0 (58.0)	26.0 (58.0)	34.9 (38.0)	0.220
EQ-5D-5 L, scored from < 0 to 1 (median, IQR)	0.615 (0.789)	0.675 (0.789)	0.507 (0.685)	0.085
EQ-5D-5 L (VAS), scored from 0 to 100 (median, IQR)	60 (80)	60 (70)	50 (70)	0.124
PCS, scored from 0 to 52 (median, IQR)	24 (47)	18 (38)	26.50 (42)	0.097
PCS M, scored from 0 to 12 (median, IQR)	4 (12)	4 (9)	4.50 (11)	0.075
PCS R, scored from 0 to 16 (median, IQR)	8 (16)	6 (12)	9 (13)	0.098
PCS H, scored from 0 to 24 (median, IQR)	12 (22)	8 (18)	12 (20)	0.084

EQ-5D-5 L Euro Quality of Life 5-Dimensions and 5-Levels; IQR Interquartile Range; ODI Oswestry Disability Index; PCS Pain Catastrophising Scale; PCS M Magnification subscale of the PCS; PCS R Rumination subscale of the PCS; PCS H Helplessness subscale of the PCS; SD Standard Deviation; VAS Visual Analogue Scale

Feasibility outcomes

Of the 227 assessed patients diagnosed with non-specific CLBP, 150 (66%) were included in the program, and 77 (44%) were excluded (Fig. 2). The reasons for exclusion were varied. There was the unwillingness to participate in the program (49%, 38/77), having another type of pain predominant instead of non-specific CLBP like neck, abdominal, inguinal or radicular pain (30%, 23/77), having difficulties in reaching the facilities where the program conducted (12%, 9/77) and not being able to speak and read Spanish or Catalan (5%, 4/77). Furthermore, there was a diagnosis of a mental health disease like schizophrenia (4%, 3/77). Full treatment completion was achieved with 121 participants (80.7%). With that, 29 patients were unable to complete the program within a 4-month period (19.3%). The main reasons for not completing the treatment were a lack of availability (48%, 14/29) or not wanting to continue with the program for different reasons like problems getting to the facilities, or even because they felt that the program did not meet their needs as they saw them (52%, 15/29). No harm was done nor were any side-effects detected throughout the trial.

Clinical outcomes

The second recruitment wave was between January 2023 and May 2023. During this period, 61 patients were assessed for eligibility, and 41 (67.2%) were included in the study. Exclusion and follow-up details are presented in Fig. 2.

Baseline descriptive statistics revealed no differences between the groups at baseline in any outcome measure (Table 1). The between groups analysis showed no significant differences at the baseline and post-treatment assessment points. However, significant differences and a medium size effect were observed between groups in VAS pain (p = .017, r = .408) at the 4-month follow-up (Table 2). The intragroup analysis revealed no significant variations over time in the control group. In the intervention group, VAS Pain (p = < 0.001) and VAS-QoL (p = .030) significantly improved over time. Detailed intragroup comparison results are presented in Table 3.

Table 2.

Detailed between-groups comparison results

Variable	Baseline a			Post-treatment (2 months) b			4-month follow-up c			Effect size r
	Control Group (n = 21)	PAINDOC Group (n = 20)	Mann Whitney’s U (p-value)	Control Group (n = 19)	PAINDOC Group (n = 17)	Mann Whitney’s U (p-value)	Control Group (n = 19)	PAINDOC Group (n = 15)	Mann Whitney’s U (p-value)
VAS pain, scored from 0 to 10 (median, IQR)	6.5 (5.0)	7.5 (2.75)	0.304	6.5 (6)	6 (8)	0.232	7 (6.5)	5 (8)	0.017	.161a, .205b, .408c
ODI, scored from 0 to 100 (median, IQR)	26.0 (58.0)	34.9 (38.0)	0.220	26 (52)	34 (44)	0.552	31 (50)	26 (58)	0.691	.192a, .101b, .069c
EQ-5D-5 L, scored from < 0 to 1 (median, IQR)	0.675 (0.789)	0.507 (0.685)	0.085	0.694 (0.964)	0.675 (0.639)	0.827	0.671 (684)	0.675 (0.744)	0.828	.269a, .040b, .037c
EQ-5D-5 L (VAS), scored from 0 to 100 (median, IQR)	60 (70)	50 (70)	0.124	60 (60)	60 (60)	0.900	60 (70)	55 (40)	0.811	.240a, .022b, .042c
PCS, scored from 0 to 52 (median, IQR)	18 (38)	26.50 (42)	0.097	16 (40)	25 (33)	0.093	14,50 (47)	19 (38)	0.302	.259a, 283b, .180c
PCS M, scored from 0 to 12 (median, IQR)	4 (9)	4.50 (11)	0.075	3 (9)	6 (8)	0.061	3 (11)	5 (8)	0.104	.278a, .318b, .283c
PCS R, scored from 0 to 16 (median, IQR)	6 (12)	9 (13)	0.098	5 (15)	8 (15)	0.156	4,50 (15)	7 (12)	0.309	.258a, .239b, .177c
PCS H, scored from 0 to 24 (median, IQR)	8 (18)	12 (20)	0.084	8 (16)	10 (15)	0.087	6,50 (22)	10 (18)	0.574	.269a, .286b, .100c

^a Baseline; ^b Post-treatment (2 months); ^c 4 months follow-up; EQ-5D-5 L Euro Quality of Life 5-Dimensions and 5-Levels; IQR Interquartile Range; ODI Oswestry Disability Index; PCS Pain Catastrophizing Scale; PCS M Magnification subscale of the PCS; PCS R Rumination subscale of the PCS; PCS H Helplessness subscale of the PCS; VAS Visual Analogue Scale

Table 3.

Detailed intragroup comparison results

Variable	Groups	Baseline vs. post-treatment (p-value) a	Baseline vs. 4-month follow-up (p-value) b	Post-treatment vs. 4-month follow-up (p-value) c	Friedmann’s ANOVA	Effect size r
VAS pain, scored from 0 to 10	Control	0.985	0.887	1.000	0.717	.095a, .049b, .195c
	PAINDOC	0.001	0.002	0.147	< 0.001	.439a, .513b, .088c
ODI, scored from 0 to 100	Control	0.526	0.500	0.759	0.904	.146a, .159b, .072c
	PAINDOC	0.063	0.068	0.172	0.076	.420a, 471b, .352c
EQ-5D-5 L, scored from < 0 to 1	Control	0.679	0.836	0.408	0.526	.107a, .049b, .195c
	PAINDOC	0.070	0.047	0.733	0.193	.439a, .513b, .088c
EQ-5D-5 L (VAS), scored from 0 to 100	Control	0.458	0.053	0.299	0.343	.170a, .456b, .245c
	PAINDOC	0.009	0.056	0.341	0.030	.630a, .493b, .246c
PCS, scored from 0 to 52	Control	0.793	0.687	0.660	0.901	.060a, .095b, .104c
	PAINDOC	0.064	0.030	0.729	0.188	.449a, .560b, .089c
PCS M, scored from 0 to 12	Control	0.922	0.729	0.719	0.882	.023a, .082b, .085c
	PAINDOC	0.420	0.275	0.873	0.771	.195a, .282b, .041c
PCS R, scored from 0 to 16	Control	0.641	0.756	0.566	0.874	.107a, .073b, .135c
	PAINDOC	0.073	0.107	0.707	0.097	.435a, .416b, .097c
PCS H, scored from 0 to 24	Control	0.755	0.697	0.588	0.611	.072a, .092b, .128c
	PAINDOC	0.027	0.032	0.239	0.061	.535a, .519b, .285c

^a Baseline vs. post-treatment (2 months); ^b Baseline vs. 4 months follow-up; ^c Post-treatment (2 months) vs. 4 months follow-up; EQ-5D-5 L Euro Quality of Life 5-Dimensions and 5-Levels; ODI Oswestry Disability Index; PCS Pain Catastrophizing Scale; PCS M Magnification subscale of the PCS; PCS R Rumination subscale of the PCS; PCS H Helplessness subscale of the PCS; VAS Visual Analogue Scale

Full trial sample estimation

Sample size estimation was performed with the G*Power3 software and pain intensity using a VAS was used as a primary outcome measure. The effect size (d) of the intervention group relative to the control group was calculated based on the 4-month follow-up VAS pain score (means and standard deviations: 6.63 ± 1.56, control group; 4.87 ± 2.19, intervention group). Thus, with a d = 0.929, a power of 0.95, and an α level of 0.05, it was estimated that 26 participants would be needed for each group (a total sample size of 52). Therefore, accounting for a 20% drop-out rate, the full trial enrolment goal was set to 31 participants per group (a total sample size of 62 subjects).

Discussion

The present study aimed to assess the feasibility and appropriateness of the multidisciplinary treatment program named Comprehensive Care Program for People with Chronic Pain (PAINDOC) as a therapeutic option for patients with CLBP. Secondarily, the objective included evaluating the potential effect of the program in terms of lessening pain intensity, PC and CLBP-related disability as well as improving QoL when compared with the pharmacological control group. The findings show that the PAINDOC program is a feasible treatment. Moreover, the exploratory results of the study suggest that it could be an effective treatment in reducing pain intensity and self-reported QoL.

Regarding the treatment of CLBP, although clinical guidelines recommend multidisciplinary treatment based on the biopsychosocial model of pain as the first option [41–43], it is not widely applied in clinical practice. Our study is in line with previous work that has already explored the feasibility of multidisciplinary programs for CP treatment.

Nishie et al. performed a program of 6 sessions, which is comparable to our 8 sessions. It obtained a remarkably similar completion rate (80%) [44]. In contrast to our program, their sessions were individual. In this regard, it is worth noting that group sessions can optimize resources [45] and produce therapeutic effects [45, 46] (e.g., interpersonal learning, group cohesion, altruism). In fact, this format is recommended over individual sessions for non-specific CLBP [43, 47].

In the same line, Monticone et al. evaluated a 2-month individual multidisciplinary rehabilitation program consisting of 16 sessions of TE and 8 sessions of CBT [48]. Interestingly, they obtained a 100% completion rate. It might be related to the small sample size (N = 20) and the €230 compensation offered to patients who completed the program by the Italian healthcare system.

Remarkably, most of the participants in the previously mentioned programs were not working (60%, 81.25%) [44, 48], which is consistent with our study population (70.73%). Moreover, the high presence of unemployed people is concordant with the fact that low-back pain is the leading cause of global productivity loss [2] and the most common cause of disability worldwide [42]. Importantly, the completion rate might be influenced by the availability and flexibility in terms of time of unemployed individuals to attend treatment sessions.

In this regard, some working groups have designed programs with online sessions to increase the accessibility of this type of treatment programs (e.g. to people with less availability or who live in areas far from the treatment center). Garreta-Catala et al. carried out an online multidisciplinary intervention that integrated physical rehabilitation and psychosocial intervention in 8 sessions [49]. Despite their similar design (weekly sessions and multidisciplinary working group), with the online approach, they only obtained a program session completion rate of 67%. Furthermore, it is not clear whether they were more accessible for employed people than the unemployed. Although they had a higher percentage of employed participants than in other studies (50%), they had only included working-age participants.

In our case, the face-to-face format of the PAINDOC program seems to have proven itself accessible. Among all the participants evaluated in the first recruitment wave, none refused to participate due to incompatibility with work and only 8.2% (5/61) did so in the second wave. In addition, only 4% (9/227) of the first wave and 3.3% of the second wave (2/61) refused to participate due to difficulty getting to the facilities. Moreover, it is important to reduce participant burden in studies and treatment programs through both the design of the intervention itself and through simplified data collection processes [50]. Notably, the PAINDOC program has synthesized therapeutic educational content based on CBT (Empowered Relief^®), PNE, mindfulness meditation, CBT-based psychotherapy, and TE in 8 sessions. In this way, participants undertake only 8 sessions, held weekly, with flexibility for schedule changes to facilitate attendance. This schedule of a few short, well-spaced sessions makes it possible for participants to reconcile their personal and work life with the program. Research in multidisciplinary treatment programs should aim not only at defining the most effective combination of therapies, but also at determining the optimal number of sessions (resources used) to obtain the expected benefits for patients with the minimum burden for them (value-based healthcare) [51]. In this sense, shorter treatment programs like Empowered Relief^® (single 2-hour session) can have similar clinical effects to full CBT programs, but with much higher completion rates [17].

To conclude, the multidisciplinary PAINDOC program is a feasible treatment. Furthermore, the results showed a reduction in pain intensity in the intervention group at the 4-month follow-up. In addition, there was also a trend towards improvement in CLBP-related disability, PC and QoL, with statistically significant differences in pain intensity and QoL in the PAINDOC group at 4 months. Our data confirm the feasibility and suggest a possible effectiveness of the PAINDOC program as a multidisciplinary intervention for patients with non-specific CLBP, although a randomized clinical trial is needed to determine the clinical effect of the program.

It is essential to develop well-coordinated, accessible, and simply structured multidisciplinary programs to improve adherence to them. They would lead to better clinical outcomes, better QoL and less need for pharmacological or invasive intervention. It would be valuable to explore, through a qualitative research approach, the acceptability of and detection of barriers to the program by patients receiving this type of intervention. It would also be worthwhile to evaluate the long-term clinical impact of the program and the possible decrease in analgesic consumption.

Limitations

Several limitations of our study should be mentioned. One is that our cohort included a low percentage of working people, a group of patients where engagement may be lower, so the findings could not be generalized to this population subgroup. In this regard, to make it more accessible to younger and employed people, not only could more afternoon sessions be offered to facilitate after-work attendance, but also hybrid programs, so that some sessions could be attended from home. In addition, participants placed in the intervention group received a greater degree of attention from healthcare professionals, which could lead to a Hawthorne effect or intensify the placebo effect. This may enhance patient satisfaction and the expectation of improvement with the treatment received in the intervention group. Moreover, the very nature of the intervention and the impossibility of including a placebo group must be considered a limitation. In a future randomized clinical trial, an option to consider would be the introduction of some follow-up visits in the control group performing usual care, in order to balance both the care received between groups and the placebo effect. Furthermore, to compensate for the lack of a placebo group, it might also be interesting to consider giving written recommendations related to the disciplines worked on in the PAINDOC program to the control group participants. Regarding the statistical analysis, multiple comparisons could induce a type I error, so it would be advisable for a future clinical trial to apply the Bonferroni correction to adjust for this error. Finally, there is the fact that this is a single-center study with a small sample size, so external validity is limited.

Conclusions

This study showed that the PAINDOC multidisciplinary program is a feasible treatment for patients with non-specific CLBP. Furthermore, the exploratory results of this study suggest that it could be an effective treatment to reduce pain intensity and improve on self-reported QoL in these patients, although a randomized clinical trial is needed to prove its effectiveness. In addition, such interventions could also be beneficial to reduce CLBP-related disability and PC. Future studies should be directed toward determining whether the PAINDOC program can bring about such improvements.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

CBT

Cognitive behavioral therapy

CP

Chronic pain

CLBP

Chronic low back pain

EQ

5D–5 L–5–dimension, 5–level, self–administered euro quality of life questionnaire

ODI

Oswestry disability index

PAINDOC

Comprehensive care program for people with chronic pain (programa d’Atenció integral per a persones amb dolor Crònic)

PC

Pain catastrophizing

PCS

Pain catastrophizing scale

PNE

Pain neuroscience education

QoL

Quality of Life

SPSS

Statistical package for the social sciences

TE

Therapeutic exercise

VAS

Visual analogue scale

WHO

World health organization

Author contributions

All authors contributed to the conception and design of the study. Christian Dürsteler (CD), Marc Terradas-Monllor (MT-M) and Anna Dalmau-Roig (AD-R) provided the idea for the study, established the hypothesis, and wrote the original proposal. AD-R, MT-M, Mirari Ochandorena-Acha and CD significantly contributed to the drafting of this paper. All the authors were involved in critically revising the manuscript. This paper was written by AD-R, MT-M and CD with input from all the co-authors.

Funding

This research received no specific grant from any funding agency in the public, commercial, or non-profit sectors.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The study protocol was approved by the Research Ethics Committee of the Hospital Clínic de Barcelona (HCB/2023/0177). All patients included in this trial gave written informed consent to participate. This clinical trial was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice Guidelines (ICH E6 R2). Patient confidentiality was guaranteed in accordance with current Spanish legislation (LOPD 3/2018). No human tissue samples were collected, used, or stored in this clinical trial.

Consent for publication

Not applicable. This manuscript does not contain information or images that could lead to the identification of a study participant.

Competing interests

The authors declare no competing interests.