The burden associated with neuropathic pain in Western Europe

Abstract

Background:

The purpose of this study is to assess the burden of neuropathic pain (NeP) on health-related quality-of-life (HRQoL), health status, employment status, absenteeism and presenteeism, and direct medical costs in Western Europe.

Methods:

Data are from the 2010 National Health and Wellness Survey (NHWS) for five countries in western Europe: the UK, France, Spain, Germany, and Italy. Among subjects who reported experiencing pain in the past month, those who attributed their pain to NeP were compared with those who attributed their pain to another chronic pain condition other than NeP (the latter was the reference group). These two groups were compared on demographic and both pain and non-pain related comorbidities. Generalized linear models were used to estimate the independent contribution of the presence of NeP on: (a) HRQoL (using the SF-12v2); (b) self-reported health status (the first item of the SF-12v2); (c) employment status; (d) absenteeism and presenteeism (using the WPAI questionnaire); and (e) direct medical costs (estimated from self-reported healthcare resource use and unit costs from the literature).

Results:

Relative to the chronic pain reference group, subjects with NeP reported a higher prevalence of severe daily pain (38.12% vs 12.67%, p < 0.05), lower labor force participation (39.68% vs 55.56%; p < 0.05), higher prevalence of sleep difficulties (59.14% vs 46.73%; p < 0.05), insomnia (45.61% vs 29.78%; p < 0.05) anxiety (42.42% vs 31.99%; p < 0.05), and depression (35.25% vs 24.03%; p < 0.05). NeP subjects reported higher rates of absenteeism (39.78% vs 21.47%; p < 0.05) and presenteeism (86.48% vs. 66.70%; p < 0.050). Direct medical costs were approximately twice as high compared to non-NeP controls. In addition, >80% of NeP patients reported having other pain conditions. Regression results amplified these findings by indicating the independent contribution of confounding factors on the presence of NeP.

Limitations:

The NHWS is an Internet-based survey and may not be representative of the respective country populations if Internet access is limited. Second, respondents are asked to report their experience of pain. Although respondents are asked if their pain condition has been diagnosed by a physician there is no separate clinical confirmation of the presence of pain, pain conditions reported, and the presence of comorbidities.

Conclusions:

The presence of NeP is associated with an increased disease burden in the chronic pain population. This is seen in terms of HRQoL, health status, employment experience, and direct medical costs.

Keywords::

• Burden
• Employment status
• Health outcomes
• Neuropathic pain

Introduction

The International Association for the Study of Pain (IASP) defines neuropathic pain (NeP) as ‘pain caused by a lesion or disease of the somatosensory system’1. Patients with NeP are heterogeneous, presenting with a variety of sensory symptoms and pain qualities2. NeP is to be distinguished from nociceptive or inflammatory pain, although in many cases both nociceptive and neuropathic components may contribute to pain. In the case of chronic low back pain, for example, epidemiological studies suggest that up to 20–40% of patients suffer from pain that may have a NeP component3–7.

There are few quantitative assessments of the contribution of NeP to the overall burden of pain. Although the number of studies is limited, and often restricted to NeP populations, they point to the potentially significant effect of NeP on medication and healthcare utilization and a corresponding negative health-related quality-of-life8–11. A review by O’Connor12, for example, provides a useful summary of evidence to date for the impact of NeP on HRQoL and, more recently, a meta-analysis of health utilities by Doth et al.13 A recent retrospective points to the relatively high levels of healthcare resources in a UK general practice population with painful neuropathic disorders14. Even so, as far as can be determined, there are no studies that have addressed the question of the impact of NeP on HRQoL, health status, employment status, absenteeism, and presenteeism, as well as direct medical costs from a pan-European perspective.

While individual country studies have considered the association of NeP with direct medical costs and reduced productivity15 as well as profiling characteristics against matched non-pain populations8,14,16, it is unclear as to whether claims from these studies translate to other geographies, or whether they capture all potential attributes of pain experience and comorbidities. In particular, population studies have pointed to the association between NeP and psychiatric conditions—in particular the relationship between NeP and depression17.

The present study focuses on both the characteristics of the NeP population, to include the experience of multiple pain conditions and comorbidities, but also on the independent contribution of the presence of chronic pain to the burden of illness experienced by the chronic pain population in the UK, France, Spain, Germany, and Italy. The impact of the burden of pain is considered in terms of (i) health-related quality-of-life (HRQoL); (ii) self-reported health status; (iii) labor force participation, absenteeism, and presenteeism; and (iv) direct medical costs. The present analysis presents, for the first time, estimates of the impact of NeP based upon a population health survey that covers all five countries. The question of whether country of residence contributes to these outcomes is also addressed.

Methods

The neuropathic pain population

This study utilizes data from the 2010 National Health and Wellness Survey (NHWS) of Europe18. The NHWS reports on the health status and health experience of a representative quota sample (n = 57,800) of the population 18 years of age and over France, Germany, Italy, Spain, and the UK. Other countries included in NHWS surveys are the US, Brazil, Russia, urban China, and Japan. The NHWS is unique as a national population-based health survey in the range of health conditions covered, respondent socio-economic profile, health status and outcomes, and utilization of pharmaceuticals. The NHWS is Internet-based and has been used on a number of occasions to assess the association of chronic pain severity and frequency on HRQoL, health utilities, labor force status, absenteeism and presenteeism, and direct medical costs for the five countries reported here19–23.

The present analysis is based upon all respondents to the NHWS in 2010 who were considered to have experienced chronic pain (20.17% of the adult population). Respondents were asked if they had experienced pain in the last 30 days and to indicate, from a pre-assigned list, which pain conditions they had experienced. Persons reporting only acute pain conditions—headache, menstrual pain, dental pain, or migraine—were excluded. The chronic pain group was then divided into those who reported NeP (n = 512) and those who did not; the latter is the control group for the analysis. Respondents also detailed their socio-demographic characteristics, other health conditions experienced in the last 12 months, health risk behaviors (smoking, alcohol use, height, and weight), employment status, and healthcare resource use in the previous 6 months.

Health-related quality-of-life (HRQoL)

Respondents completed the 4-week version of the SF-12 generic HRQoL instrument24. The SF-12 yields physical component scores (PCS) and mental component scores (MCS). A sub-set of the SF-12 questionnaire items generates the SF-6D absolute utility score25,26. The first item of the SF-12 questionnaire asks respondents to report their general health status on a five category scale: excellent, very good, good, fair, poor. Each of these elements of the SF-12 is used in the current analysis. The SF-6D score is a preference-based index measure of health. The preference scores are derived from a UK population sample where individuals are asked to value a selection of health states defined by a sub-set of items in the SF-12 questionnaire. The maximum score is unity (‘perfect health’), with lower scores representing progressively worse health experience.

Employment status, absenteeism, and presenteeism

As well as reporting current employment status (employed full-time, employed part-time, self-employed, unemployed, and not in the labor force) the NHWS uses the WPAI scale to measure the impact of health status on employment-related activities. The WPAI questionnaire measures work time missed and work and activity impairment because of a specified health problem during the past 7 days27. Two measures are used in the present study: (i) absenteeism, the percentage of work time missed in the past 7 days; and (ii) presenteeism, the extent to which productivity at work was impaired.

Healthcare resource utilization

The 2010 NHWS also asks respondents about their use of healthcare resources in the past 6 months. Resource utilization is considered in terms of visits or events as they relate to: (i) emergency room visits in the last 6 months; (ii) number of times hospitalized in the last 6 months; and (iii) number of visits to traditional healthcare providers in the last 6 months. The number of visits reported by country for each of these resource elements is translated to direct medical costs by applying the average cost for resource event for each country; aggregating for each respondent to yield an estimate of individual direct medical costs. Respondents are dropped from the analysis if there is no resource units reported.

Analysis overview

The part consists of two parts. The first part describes the experience of pain; the second part presents a series of multivariate models to assess the independent effect of the presence of neuropathic pain, controlling for the presence of potential confounding factors on outcomes.

The first part of the analysis considers the following attributes of pain experience for those experiencing NeP (in association with other pain conditions) and controls:

• severity and frequency of pain experienced,

• characteristics of those reporting NeP,

• other chronic pain conditions reported,

• major comorbidities,

• HRQoL (SF-12 PCS and MCS; SF-6D utilities),

• health status,

• employment status,

• absenteeism and presenteeism, and

• direct medical costs.

The second part of the analysis uses a number of generalized linear regression models with the expectation that the presence of NeP, relative to the control population, will be associated with deficits in HRQoL, health status, and employment experience, together with an increase in direct medical costs. The covariates included in the regression models are:

• Socio-demographic characteristics (age, gender, household income, employment status, and education);

• Respondent health-risk behaviors (body mass index [BMI], alcohol use, smoking);

• Presence of physical comorbidities (Charlson Comorbidity Index (CCI))28; and

• Country effects (Germany as reference country).

The Charlson Comorbidity Index (CCI) was originally designed as a measure of the risk of 1-year mortality associated with comorbidity. The CCI identifies 19 categories of comorbidity (which are identified in the NHWS). Each category is assigned a weight such that the overall score reflects the cumulative increased likelihood of 1-year mortality.

The choice of generalized linear model is determined by the characteristics of the dependent or outcome variable. The estimators are (i) ordinary least squares for HRQoL (PCS and MCS) and SF-6D utilities; (ii) semi-logarithmic ordinary least squares for healthcare cost; (iii) ordered logit for health status, absenteeism, and presenteeism; and (iv) binary logit for labor force status. Although the descriptive tables are weighted, to reflect differences in sampling ratios, the multivariate results are presented in unweighted form; these estimates are preferred where sampling weights are solely a function of independent variables included in the model because the estimates are unbiased, consistent, and have smaller standard errors than weighted OLS estimates29.

Results

Prevalence of Pain and Neuropathic Pain

Overall, as a weighted estimate from the 2010 NHWS, 2.09 million persons in France, Germany, Italy, Spain, and the UK reported experiencing NeP in the last 30 days. This represents 4.00% of the total chronic pain population of 51.18 million in these five countries. Compared to the control group, those reporting NeP experienced more severe and frequent pain (Table 1). In the NeP group, 41.18% reported severe pain, compared with only 15.92% in the reference group (p < 0.05). Severe daily pain was reported by 38.12% in the NeP group compared with only 12.67% (p < 0.05) in the control group. The difference is less marked for those reporting moderate daily pain (32.9% vs 27.7%; p < 0.05).

Table 1.  Severity and frequency of pain.

Pain severity and frequency	Persons reporting neuropathic pain (%)	Persons not reporting neuropathic pain (%)
Mild	8.31	23.68
Moderate	50.51	60.40
Severe	41.18	15.92
Daily pain	75.04	44.45
4–6 times a week	9.73	11.57
2–3 times a week	9.41	18.03
Weekly or less	5.83	25.95
Mild daily pain	4.00	4.10
Moderate daily pain	32.92	27.68
Severe daily pain	38.12	12.67
Less than daily	24.96	55.55

Note: Estimates weighted by age and gender for each country and combined for the UK, France, Spain, Germany and Italy. Source: NHWS, 2010.

Characteristics of the pain populations

Persons reporting NeP tended to be older than those in the control group, although differences in terms of gender, education, and health risk factors were not statistically significant (Table 2). Persons reporting NeP had a significantly lower labor force participation than the controls (39.7% vs 55.6%; p < 0.05). At the same time, a significantly greater proportion of respondents in the NeP group reported household income less than €20,000 (39.5% vs 31.8%, p < 0.05) compared to controls.

Table 2.  Respondent characteristics.

Respondent characteristics		Persons reporting neuropathic pain (%)	Persons not reporting neuropathic pain (%)
Age	18-39	14.86	27.43
	40-59	47.33	41.80
	60 and over	37.81	30.77
Gender	Female	59.73	59.75
	Male	40.27	40.25
Education level	Less than High School	14.92	14.84
	High School	42.74	48.23
	University	42.34	36.93
Annual household income	Under €20,000	39.50	31.84
	€20,000–49,999	38.93	44.21
	€50,000 or greater	9.99	11.27
	Declined to answer	11.58	12.68
Employment Status	Employed full time	18.96	33.17
	Self-employed	7.75	5.52
	Employed part time	8.85	10.31
	Unemployed	2.86	6.56
	In work force	39.68	55.56
	Not in work force	60.32	44.42
BMI	Underweight	1.74	2.17
	Normal	22.81	30.80
	Overweight	38.68	35.93
	Obese	28.77	24.66
	Morbidly Obese	5.15	3.91
	Not estimated	2.84	2.53
	Current smoker	25.45	23.84
	Use alcohol	8.03	8.25
Charlson comorbidity Index	0	50.99	70.14
	1	27.69	19.76
	2	8.60	6.68
	3	6.74	2.10
	4	3.01	0.70
	5 or more	2.97	0.63
Country	UK	34.80	27.51
	France	16.05	16.80
	Spain	10.39	11.99
	Germany	29.35	32.13
	Italy	9.41	11.56

Note: Estimates weighted by age and gender for each country and combined for the UK, France, Spain, Germany and Italy. Source: NHWS, 2010.

Chronic pain conditions reported by respondents are detailed in Table 3. For those reporting neuropathic pain relative to controls, the most prevalent pain conditions reported were back pain (59.30% vs 63.58%; p < 0.05); joint pain (57.45% vs 47.46%, p < 0.05); shoulder pain or stiffness (40.96% vs 30.36%, p < 0.05) and neck pain (35.48% vs 29.48%, p < 0.05). Once the presence of pain conditions overall is assessed, 82.56% of all those with neuropathic pain reported at least one other pain condition.

Table 3.  Pain conditions reported.

Pain conditions reported	Persons reporting neuropathic pain (%)	Persons not reporting neuropathic pain (%)
Back pain	59.30	63.58
Joint pain	57.45	47.46
Shoulder pain/stiffness	40.96	30.36
Neck pain	35.48	29.48
Headache	31.27	30.34
Arthritis pain	24.54	19.04
Migraine	17.04	13.54
Fibromyalgia pain	13.30	5.30
Dental problems	12.48	11.53
Surgery or medical procedure pain	11.84	6.48
Other	7.99	7.27
Sprains or strains pain	6.57	8.64
Menstrual cycle	5.93	9.23
Broken bone pain	4.08	3.92
Cancer pain	1.30	0.86
Post herpetic neuralgia pain	1.27	0.27
Persons only reporting neuropathic pain	17.44	n/a

Note: Estimates weighted by age and gender for each country and combined for the UK, France, Spain, Germany and Italy. Source: NHWS, 2010.

As well as reporting multiple pain conditions, the majority of respondents with pain also reported one or more comorbidities (Table 4). The most commonly reported comorbidities were sleep difficulties, headache, and insomnia (respondents could report both sleep difficulties and insomnia). This holds for both those reporting NeP and controls. The prevalence of comorbidities was significantly higher than for controls for virtually all comorbidities identified: sleep difficulties 59.14% vs 46.73%; (p < 0.05); insomnia 45.61% vs 29.78% (p < 0.05); anxiety 42.42% vs 31.99% (p < 0.05); high blood pressure 38.44% vs 28.88% (p < 0.05); and depression 35.25% vs 24.03% (p < 0.05). Only headache, alone among the more commonly reported comorbidities, had a significantly higher prevalence in the control group (54.08% vs 60.02%, p < 0.05).

Table 4.  Major comorbidities.

Major comorbidities	Persons reporting neuropathic pain (%)	Persons not reporting neuropathic pain (%)
Sleep difficulties	59.14	46.73
Headache	54.08	60.02
Insomnia	45.61	29.78
Anxiety (not otherwise specified)	42.42	31.99
High blood pressure (Hypertension)	38.44	28.88
Depression	35.25	24.03
Heartburn	32.83	33.94
Abdominal bloating	32.72	26.28
Migraine	31.02	29.10
Abdominal pain	30.87	27.78
High cholesterol	27.80	22.89
Arthritis	26.54	21.43
Diabetes (Type 1 or Type 2)	24.97	8.14
Restless Legs Syndrome	22.93	10.59
Irritable bowel syndrome	19.80	10.64
GERD or acid reflux	19.20	11.08
Social anxiety disorder	10.24	6.19
Generalized anxiety disorder	9.90	6.09
Panic disorder	8.51	5.11

Note: Estimates weighted by age and gender for each country and combined for the UK, France, Spain, Germany and Italy. Source: NHWS, 2010.

Health-related quality-of-life and health status

Results for HRQoL and health status are presented in Table 5. The SF-12 PCS was significantly greater among those with NeP than for controls (31.29 vs 41.50; p < 0.05). The difference in SF-12 MSC coefficients was not statistically significant. (This result is discussed in more detail later as it is in contrast to the multivariate modeled results.) Estimated SF-6D utilities pointed to a marked difference in HRQoL experience. In the NeP group, as would be expected, average utility score was 0.56 compared to 0.65 (p < 0.05) for controls. Among those with NeP, 30.93% described their health as poor compared with only 8.98% (p < 0.05) in the reference group. Although less marked, those reporting NeP tended to report their health status as only fair (43.72% vs 31.33%; p < 0.05). These estimates were in addition to those reporting their health as excellent or very good (4.14% for the NeP group vs 18.93% for the control group (p < 0.05).

Table 5.  Health-related quality-of-life and health status.

Pain conditions reported	Persons reporting neuropathic pain	Persons not reporting neuropathic pain
SF-12 PCS (SD)	31.29 (11.05)	41.50 (11.00)
SF-12 MCS (SD)	40.16 (11.51)	43.90 (11.06)
SF-6D utilities (SD)	0.56 (0.11)	0.65 (0.12)
Health status (%)
Excellent	0.30	2.29
Very good	3.84	16.64
Good	21.21	40.77
Fair	43.72	31.33
Poor	30.93	8.98
Total	100.00	100.00

Note: Estimates weighted by age and gender for each country and combined for the UK, France, Spain, Germany and Italy. Source: NHWS, 2010.

Employment status, absenteeism, and presenteeism

The association of NeP with employment status, absenteeism, and presenteeism is detailed in Table 6. Subjects with NeP reported a significantly lower level of labor force participation (39.68% vs 55.56%; p < 0.05). At the same time, for those in employment, subjects with NeP reported greater absenteeism (39.78% vs 21.47%; p < 0.05) and a higher incidence of health problems leading to presenteeism (86.48% vs 66.70%; p < 0.05).

Table 6.  Employment status, absenteeism, and presenteeism.

Employment status		Persons reporting neuropathic pain (%)	Persons not reporting neuropathic pain (%)
In work force		39.68	55.56
Not in work force		60.32	44.44
Workplace absenteeism	None	60.22	78.53
	Low (1–10%)	9.22	5.71
	Medium (11–25%)	5.26	5.25
	High (26–100%)	25.31	10.51
Workplace presenteeism	No impact	13.52	33.30
	Low impact	7.67	12.96
	Medium impact	17.72	22.64
	High impact	61.09	31.09

Note: Estimates weighted by age and gender for each country and combined for the UK, France, Spain, Germany and Italy. Source: NHWS, 2010.

Direct medical costs

Two aspects of the impact of the presence of NeP on healthcare resources were considered: direct medical costs and traditional provider visits (Table 7)30. Overall, taking into account traditional provider visits, emergency room visits, and hospitalizations, the mean medical costs incurred, for those reporting costs, in the 6-months prior to the NHWS response, were €788.18 for those with NeP and €414.66 (p < 0.05) for controls. In the case of traditional healthcare visits, the difference was less marked (€356.89 vs €205.97; p > 0.05). Given the skewed nature of the cost distribution median costs results are presented. These costs were €350.16 for those with NeP and €146.70 for controls (p > 0.05).

Table 7.  Direct medical costs and resource units.

Direct medical costs (6 months)	Persons reporting neuropathic pain (€; SD)	Persons not reporting neuropathic pain (€, SD)
Mean cost (SD)
Traditional provider visits	356.89 (357.29)	205.97 (243.46)
Emergency room visits	58.94 (182.33)	30.71 (110.66)
Hospitalizations	372.35 (969.59)	177.98 (1075.20)
Total direct cost	788.18 (1164.28)	414.66 (1173.75)
Median cost (25th–75th quartile)
Traditional provider visits	240.00 (125.02–466.88)	140.00 (58.68–262.62)
Total direct cost	350.16 (142.88–968.22)	146.70 (60.00–362.98)

Note: Estimates weighted by age and gender for each country and combined for the UK, France, Spain, Germany and Italy; costs assigned are: UK, traditional provider visit €29.18, emergency room visit €109.42, hospitalization €668.79; France, traditional provider visit €20.00, emergency room visit €23.00, hospitalization €1371.50; Germany, traditional provider visit €29.34, emergency room visit €157.60, hospitalization €1022.19; Italy, traditional provider visit €12.91, emergency room visit €61.97, hospitalization €593.31; Spain, traditional provider visit €17.86, emergency room visit €103.60, hospitalization €802.80; SD, standard deviation. Source: NHWS, 2010.

Multivariate regression analysis

Tables 8–11 present the results of the multivariate regression analysis. In Table 8 the regression results pointed to the lower HRQoL across all measures for those with NeP. The deficit impact of NeP was greater for the PCS element of the SF-12 (−8.196; 95% CI = −9.053 to −7.339) than for the MCS (−3.190; 95% CI = −4.133 to −2.248). In the PCS model the impact of NeP was greater than any of the other covariates. On the other hand, the results for MCS were mixed. The association of NeP on SF-6D utilities was also significant (−0.078; 95% CI = −0.089 to −0.068), with a deficit effect greater than any of the other covariates.

Table 8.  Multivariate regression results:: SF-12 and SF-6D health-related quality-of-life.

Independent variables	SF-12 PCS		SF-12 MCS		SF-6D utilities
	b coeff	95% CI	b coeff	95% CI	b Coeff	95% CI
Presence of neuropathic pain^a	−8.196	−9.053 to −7.339	−3.190	−4.133 to −2.248	−0.078	−0.089 to −0.068
Socio-demographic variables
Age: 40–59 years^b	−5.267	−5.684 to −4.850	1.495	1.037 to 1.954	−0.017	−0.022 to −0.012
Age: 60 years and older^b	−4.860	−5.408 to −4.312	8.066	7.464 to 8.668	0.039	0.033 to 0.046
Gender: male^c	1.361	1.002 to 1.720	1.529	1.134 to 1.923	0.023	0.019 to 0.027
Education: High school completed^d	1.222	0.688 to 1.756	1.025	0.438 to 1.613	0.016	0.009 to 0.022
Education: University or higher^d	2.109	1.556 to 2.662	1.851	1.243 to 2.459	0.027	0.020 to 0.033
Household income: €20,000–€39,9996^e	1.305	0.903 to 1.708	1.807	1.365 to 2.250	0.023	0.018 to 0.028
Household income €40,000 and above^e	2.597	1.977 to 3.216	2.698	2.017 to 3.379	0.040	0.032 to 0.047
In labour force^f	5.387	4.967 to 5.790	2.673	2.221 to 3.125	0.055	0.050 to 0.060
Health risk behaviors
BMI: Underweight^g	−1.210	−2.329 to −0.091	−2.641	−3.871 to −1.410	−0.027	−0.041 to −0.014
BMI: Overweight^g	−1.253	−1.686 to −0.821	0.112	−0.363 to 0.587	−0.006	−0.011 to −0.001
BMI: Obese^g	−3.942	−4.419 to −3.465	−0.776	−1.301 to −0.252	−0.028	−0.034 to −0.023
BMI: Morbidly obese^g	−7.346	−8.228 to −6.465	−1.634	−2.603 to −0.664	−0.055	−0.065 to −0.044
Current smoker^h	−1.256	−1.664 to −0.848	−1.328	−1.777 to −0.880	−0.016	−0.021 to −0.011
Alcohol user^i	0.664	0.013 to 1.314	−0.523	−1.238 to 0.192	0.002	−0.006 to −0.009
Comorbidity
Charlson Comorbidity Index (CCI)	−2.304	−2.494 to −2.113	−1.233	−1.443 to −1.023	−0.022	−0.025 to −0.020
Country ^j
UK	0.637	0.190 to 1.084	1.090	0.599 to 1.581	0.018	0.013 to 0.024
France	0.721	0.220 to 1.222	0.518	−0.032 to 1.069	−0.002	−0.008 to 0.004
Spain	0.770	0.051 to 1.490	0.522	−0.269 to 1.313	0.011	0.003 to 0.020
Italy	−0.684	−1.406 to 0.038	−0.627	−1.421 to 0.167	−0.016	−0.025 to −0.008
Intercept term	42.160	41.380 to 42.950	37.150	36.290 to 38.010	0.595	0.586 to 0.605
Adj R^2	0.291		0.109		0.169
n	12,326		12,326		12,326

Reference categories: ^a absence of neuropathic pain; ^b age under 40 years; ^c female; ^d less than high school completed; ^e household income under €20,000; ^f not in labor force; ^g normal weight; ^h non-smoker; ⁱ no alcohol use; ^j= Germany.

b coeff, mean change in outcome in a given category relative to reference group.

Table 9.  Multivariate regression results: Health status.

Independent variables	Odds ratio	95% CI
Presence of neuropathic pain^a	3.725	3.124–4.441
Socio-demographic variables
Age: 40–59 years^b	2.535	2.333–2.755
Age: 60 years and older^b	1.627	1.460–1.813
Gender: male^c	0.873	0.814–0.937
Education: High school completed^d	0.806	0.727–0.895
Education: University or higher^d	0.673	0.604–0.750
Household income: €20,000–€39,999 ^e	0.732	0.676–0.792
Household income €40,000 and above^e	0.561	0.497–0.633
In labor force^f	0.424	0.391–0.461
Health risk behaviors
BMI: Underweight^g	1.652	1.324–2.063
BMI: Overweight^g	1.262	1.160–1.373
BMI: Obese^g	2.126	1.936–2.334
BMI: Morbidly obese^g	3.067	2.582–3.643
Current smoker^h	1.399	1.292–1.515
Alcohol user^i	1.028	0.907–1.166
Comorbidity
Charlson Comorbidity Index (CCI)	1.661	1.593–1.731
Country^j
UK	0.676	0.620–0.737
France	0.771	0.700–0.849
Spain	1.519	1.317–1.751
Italy	1.046	0.911–1.202
Log-likelihood ratio	−14,924.0
Pseudo R^2	0.105
n	12,326

Reference categories: ^a absence of neuropathic pain; ^b age under 40 years; ^c female; ^d less than high school completed; ^e household income under €20,000; ^f not in labor force; ^g normal weight; ^h non-smoker; ⁱ no alcohol use; ^j Germany.

Table 10.  Multivariate regression results: Labor force status, absenteeism, and presenteeism.

Independent variables	Labor force status		Absenteeism		Presenteeism
	Odds ratio	95% CI	Odds ratio	95% CI	Odds ratio	95% CI
Presence of neuropathic pain^a	0.534	0.432–0.660	2.471	1.775–3.442	2.965	2.221–3.957
Socio-demographic variables
Age: 40–59 years^b	0.896	0.809–0.993	0.826	0.709–0.963	0.812	0.733–0.899
Age: 60 years and older^b	0.064	0.056–0.073	0.844	0.627–1.136	0.666	0.545–0.814
Gender: male^c	1.568	1.428–1.721	0.847	0.731–0.981	0.843	0.765–0.928
Education: High school completed^d	1.280	1.119–1.464	0.922	0.719–1.184	0.859	0.718–1.028
Education: University or higher^d	1.883	1.635–2.167	0.800	0.619–1.033	0.760	0.634–0.910
Household income: €20,000–€39,999 ^e	2.052	1.858–2.267	0.790	0.666–0.938	0.907	0.805–1.022
Household income €40,000 and above^e	3.344	2.827–3.957	0.622	0.488–0.792	0.727	0.621–0.851
Health risk behaviors
BMI: Underweight^f	0.845	0.651–1.097	1.533	1.011–2.323	1.336	0.993–1.797
BMI: Overweight^f	1.104	0.988–1.234	1.532	1.284–1.828	1.040	0.929–1.166
BMI: Obese^f	0.950	0.842–1.072	1.523	1.247–1.859	1.217	1.069–1.387
BMI: Morbidly obese^f	0.774	0.625–0.957	2.507	1.804–3.484	1.704	1.314–2.208
Current smoker^g	1.183	1.068–1.309	1.126	0.964–1.315	0.986	0.887–1.097
Alcohol user^h	0.792	0.668–0.940	0.996	0.742–1.339	1.104	0.911–1.339
Comorbidity
Charlson Comorbidity Index (CCI)	0.797	0.758–0.837	1.321	1.228–1.420	1.381	1.296–1.472
Country^i
UK	0.720	0.644–0.807	0.672	0.558–0.810	0.696	0.616–0.787
France	1.005	0.887–1.140	0.776	0.634–0.950	0.702	0.612–0.804
Spain	1.337	1.110–1.611	1.118	0.862–1.450	0.826	0.680–1.002
Italy	1.262	1.048–1.520	0.873	0.668–1.140	1.096	0.913–1.314
Log-likelihood ratio	12,326.0		−4,346.0		−11,372.0
Pseudo R^2	0.247		0.023		0.015
n	12,326		6,071		5,721

Reference categories: ^a absence of neuropathic pain; ^b age under 40 years; ^c female; ^d less than high school completed; ^e household income under €20,000; ^f normal weight; ^g non-smoker; ^h no alcohol use; ⁱ Germany.

Table 11.  Multivariate regression results: Health costs and traditional provider visits.

Independent variables	Log of total healthcare cost		Log of provider visit cost
	b coeff	95% CI	b coeff	95% CI
Presence of neuropathic pain^a	0.518	0.414–0.622	0.401	0.323–0.480
Socio-demographic variables
Age: 40–59 years^b	0.080	0.028–0.132	0.148	0.109–0.188
Age: 60 years and older^b	−0.077	−0.144–−0.009	0.049	−0.002–0.100
Gender: male^c	−0.138	−0.183–−0.093	−0.180	−0.214–−0.146
Education: High school completed^d	−0.045	−0.112–0.021	−0.030	−0.080–0.020
Education: University or higher^d	−0.038	−0.107–0.030	−0.015	−0.067–0.037
Household income: €20,000–€39,999 ^e	−0.043	−0.093–0.008	−0.038	−0.076–0.000
Household income €40,000 and above^e	−0.094	−0.171–−0.017	−0.100	−0.158–−0.042
In Labor Force^f	−0.280	−0.331–−0.229	−0.219	−0.258–−0.181
Health risk behaviors
BMI: Underweight^g	0.136	−0.007–0.279	0.067	−0.041–0.175
BMI: Overweight^g	0.084	0.030–0.138	0.060	0.019–0.100
BMI: Obese^g	0.171	0.112–0.230	0.164	0.119–0.209
BMI: Morbidly obese^g	0.268	0.159–0.377	0.248	0.165–0.330
Current smoker^h	−0.035	−0.086–0.016	−0.017	−0.056–0.021
Alcohol user^i	0.041	−0.122–0.041	−0.056	−0.117–0.006
Comorbidity
Charlson Comorbidity Index (CCI)	0.292	0.269–0.316	0.211	0.193–0.228
Country^j
UK	−0.442	−0.497–−0.386	−0.453	−0.495–−0.411
France	−0.452	−0.514–−0.390	−0.514	−0.561–−0.467
Spain	−0.391	−0.480–−0.301	−0.603	−0.671–−0.535
Italy	−0.866	−0.954–−0.777	−0.861	−0.928–−0.794
Adjusted R^2	0.131		0.176
n	11,476		11,450

Respondents with zero costs30 are dropped from the analysis.

Reference categories: ^a absence of neuropathic pain; ^b age under 40 years; ^c female; ^d less than high school completed; ^e household income under €20,000; ^f not in labor force; ^g normal weight; ^h non-smoker; ⁱ no alcohol use; ^j Germany.

The modeled results for health status pointed to the importance of the role of NeP (Table 9). With an odds ratio (OR) of 3.725 (95% CI = 3.124–4.441), the assessed impact of NeP was significantly greater than any of the other independent variables (with the exception of the presence of morbid obesity; BMI ≥ 35 kg/m²). The country effect results are mixed; they suggest that subjects in the UK and France experienced a lower health status, while those in Spain experienced a higher health status.

The impacts of the presence of NeP on labor force participation status, absenteeism, and presenteeism are detailed in Table 10. Subjects with NeP were less likely to be in the labor force (Table 7); the modeled results indicated that, while the presence of NeP pain had a significant deficit effect (OR = 0.534; 95% CI = 0.432–0.660), other factors also contributed to reduced labor force participation (e.g., aged 40–59 years, presence of morbid obesity, CCI). Against this, there were a number of variables that had a positive association with labor force participation (e.g., household income, education). Again, the country effects were mixed. Residence in the UK was associated with a significantly lower labor force participation, but higher participation in Spain and Italy.

The association between NeP, absenteeism, and presenteeism was substantial. In the case of absenteeism (OR = 2.471; 95% CI =  1.775–3.442) this was similar to the deficit effect of morbid obesity. In the case of presenteeism the presence of NeP dominated all other covariates in its impact (OR = 2.965; 95% CI = 2.221–3.957). Country effects were mixed, with only the UK and France demonstrating that the presence of neuropathic pain was associated with higher rates of absenteeism and presenteeism.

The presence of NeP also dominated the semi-logarithmic total cost model (Table 11) with direct medical costs increasing by 51.8% (0.518; 95% CI = 0.414–0.622); the presence of morbid obesity was associated with a cost increase of 26.8%. A 1-point increase in CCI was associated with a cost increase of 29.2%. In terms of traditional provider visit costs, the presence of NeP increased these by 40% (0.401; 95% CI = 0.323–0.480).

Discussion

The results presented here point to the substantial increased burden associated with the presence of NeP when compared with the non-NeP chronic pain population. The results are consistent with those reported in previous studies which have found deficit effects on HRQoL, productivity, and healthcare resource utilization in comparison both with a non-NeP population as well as with non-pain populations8–14,17,31–34. The contribution of the present analysis is that it provides, for the first time, a comprehensive analysis utilizing the only available common pan-national health survey in Europe of the deficit association between reported NeP and HRQoL, health status, employment experience, and increased direct medical costs across five major European economies.

A key aspect of the present analysis is to point to the importance of a multivariate framework for evaluating the burden of NeP and the quantitative importance of the presence of NeP against other possible confounding factors. Simple descriptive approaches are only a first step—and may be misleading. While the deficit impact of the presence of NeP was significant in respect of the PCS, there was no significant impact (at least at the 5% level) on MCS, given the results presented in Table 5. A quite different story emerges once the multivariate results are considered (Table 8), pointing to the importance of accounting for the presence of potential confounding factors before drawing conclusions. While the impact of NeP on PCS was significantly greater than for MCS, both are significant at conventional decision levels, a result that would have been overlooked in a simple descriptive assessment. Even so, the magnitude of the impact on PCS was substantially greater than for MCS.

The quantitative association of NeP with PCS and SF-6D utilities was greater than any of the other independent variables in the model. The results for MCS were less clear cut, with BMI also contributing significantly. At the same time, age over 60 years had a major positive impact on MCS (8.066; 95% CI = 7.464–8.668) and more than compensated for the deficit impact of the presence of NeP. These multivariate results confirm those of previous studies which have utilized the NHWS to assess the impact of pain and the combination of pain severity and frequency on these dimensions of HRQoL for these five countries19 and individually for Spain21.

The presence of NeP also had a significant deficit association with self-reported health status. As detailed in Table 5, a significantly higher percentage of persons reporting NeP described their health as ‘fair’ or ‘poor’. These results are confirmed in the multivariate analysis (Table 9) where the presence of NeP had a significant deficit association (OR = 3.725; 95% CI = 3.124–4.441). These results once again confirm those for previous NHWS models of the burden of pain for these five countries19 and for Spain21.

A number of studies within NeP populations have pointed to the adverse effect the presence of this pain condition have had on employment experience12,32–34. In this analysis, the contribution of NeP was considered from two perspectives: its association with labor force status and, for those in employment, absenteeism and presenteeism (Table 10). The results are clear-cut and echo those presented in the descriptive analysis. They point to the additive role of NeP as amplifying the association of chronic pain in the absence of a neuropathic component (OR = 0.534; 95% CI = 0.432–0.660). The association of the presence of NeP on absenteeism and presenteeism was more substantial. In the case of absenteeism (OR = 2.471; 95% CI =1.775–3.442) the contribution of NeP eclipsed all other independent variables, with a similar conclusion holding in the case of presenteeism (OR = 2.965; 95% CI = 2.221–3.957).

The results presented here confirm previous assessments of the impact of NeP on healthcare utilization in the US and the UK—whether this is in terms of provider visits, emergency room visits, or hospitalizations, as well as pharmacy utilization14,33. The findings for imputed direct total medical costs and provider costs point to the magnitude of the association and the contribution of NeP within the overall costs of chronic pain.

Given the results presented here, it is worth considering the implications of pain management programs for those reporting neuropathic pain. First, it is unusual for subjects to have reported only NeP; typically subjects reported multiple other pain conditions together with a range of comorbidities. This ‘mixed pain’ syndrome is most often seen in the presence of back or joint pain—the most commonly reported pain condition. Second, subjects presenting with NeP typically reported that their overall experience of pain is both more severe and more frequent than controls. Third, subjects reporting NeP typically presented with a higher prevalence of comorbidities than controls. This is seen in terms of the CCI as well as in the prevalence of specific comorbid conditions. Particularly noteworthy are the prevalence of reported sleep difficulties and insomnia in the NeP group, together with the presence of self-reported anxiety (and associated disorders such as generalized anxiety disorder, social and panic disorder) and depression.

More appropriately perhaps, the presence of NeP, as a self-reported pain condition, should be seen as a flag or marker for multidisciplinary patient management interventions35. These interventions would be supported by the application of NeP screening tools as well as more intensive assessments of associated comorbid conditions. Unlike those not reporting NeP, the presence of this pain condition is indicative not only of more severe and frequent pain, but a more complex set of associated pain conditions, together with a significantly higher prevalence of associated co-morbidities—notably psychiatric conditions such as depression and anxiety as well as adverse sleep experience. In consequence, those reporting NeP not only have a substantially reduced HRQoL and health status, a greater likelihood of a more adverse labor market experience, and a greater likelihood of incurring higher direct medical costs.

Further research

These results amplify those of previous studies on the burden of chronic pain and the presence of NeP. As the NHWS is the only health survey that covers all five countries, the results provide a basis for not only a pan-European assessment of the presence of neuropathic pain—which may be seen as a marker for a cluster of pain conditions and high prevalence comorbidities—but also, in future analyses, in setting the stage for more disaggregated studies at the country level and for specific chronic pain conditions.

It is important to emphasize that, unlike previous studies, the present assessment is based upon a population health survey that covers five countries. Given this, the results presented here suggest a number of avenues to explore in future studies utilizing the NHWS or similar population surveys. First, the analysis can be extended to assess the quantitative contribution of pain severity and frequency to HRQoL and other outcomes within the NeP population. Second, the contribution of the presence of multiple pain conditions can be assessed present in both NeP and non-NeP populations to outcomes. Third, the impact of comorbidities in the NeP and non-NeP populations and their contribution can be studied with respect to HRQoL and other outcomes. Comorbidities of particular interest would be sleep difficulties and insomnia together with psychiatric disorders. Fourth, it is possible within the NHWS sample to undertake re-contact studies within the NeP population and to capture longitudinal data for sub-samples. These could be used to support the application of NeP screening tools to assess the claims made by respondents, as well as support alternative estimates of pain chronicity in this population. Finally, consideration should be given to extending the analysis to the country level.

Limitations

While the results presented here represent a new perspective on the experience of NeP from five countries in Western Europe, there are a number of limitations that need to be noted. First, the NHWS is an Internet-based survey and may not be representative of the respective country populations if Internet access is limited. Second, respondents are asked to report their experience of pain. Although respondents are asked if their pain condition has been diagnosed by a physician, there is no separate clinical confirmation of the presence of pain, pain conditions reported, and the presence of comorbidities.

Conclusions

The experience and burden of neuropathic pain is complex. Respondents typically presented with a number of pain conditions and comorbidities that may have been unrelated to the experience of pain. Given this, it would be a mistake to attribute the outcome deficits reported in the present study to the experience of NeP per se; rather, the presence of NeP should be considered as a flag for the association of pain experience with a range of other pain conditions and the high prevalence of certain comorbidities.

Transparency

Declaration of funding

This study was sponsored by Pfizer Inc, UK.

Declaration of financial/other interests

PC has disclosed that he is a paid consultant to Kantar Health, a company paid by Pfizer to help with the development of this manuscript. PC is also a consultant to Pfizer Inc (US) and holds stock in Pfizer Inc; CvL, JCC, and DC have disclosed that they are employees of Pfizer Inc.

Acknowledgments

PCL and DC designed the study, PCL undertook the preliminary data analysis and the drafting of the paper; CvL, JCC, and DC contributed to intellectual content by commenting and assisting in the rewriting of successive drafts as well as by writing the discussion. The peer reviewers on this manuscript have disclosed that they have no relevant financial relationships.