Avoiding Catch-22: validating the PainDETECT in a in a population of patients with chronic pain

Hans Timmerman; André P Wolff; Ewald M Bronkhorst; Oliver H G Wilder-Smith; Marcel J Schenkels; Nick T van Dasselaar; Frank J P M Huygen; Monique A H Steegers; Kris C P Vissers

doi:https://doi.org/10.1186/s12883-018-1094-4

Avoiding Catch-22: validating the PainDETECT in a in a population of patients with chronic pain

Hans Timmerman, André P Wolff, Ewald M Bronkhorst, Oliver H G Wilder-Smith, Marcel J Schenkels, Nick T van Dasselaar, Frank J P M Huygen, Monique A H Steegers, Kris C P Vissers

Anesthesiology (VUmc)

Research output: Contribution to journal › Article › Academic › peer-review

11 Citations (Scopus)

Abstract

BACKGROUND: Neuropathic pain is defined as pain caused by a lesion or disease of the somatosensory nervous system and is a major therapeutic challenge. Several screening tools have been developed to help physicians detect patients with neuropathic pain. These have typically been validated in populations pre-stratified for neuropathic pain, leading to a so called "Catch-22 situation:" "a problematic situation for which the only solution is denied by a circumstance inherent in the problem or by a rule". The validity of screening tools needs to be proven in patients with pain who were not pre-stratified on basis of the target outcome: neuropathic pain or non-neuropathic pain. This study aims to assess the validity of the Dutch PainDETECT (PainDETECT-Dlv) in a large population of patients with chronic pain.

METHODS: A cross-sectional multicentre design was used to assess PainDETECT-Dlv validity. Included where patients with low back pain radiating into the leg(s), patients with neck-shoulder-arm pain and patients with pain due to a suspected peripheral nerve damage. Patients' pain was classified as having a neuropathic pain component (yes/no) by two experienced physicians ("gold standard"). Physician opinion based on the Grading System was a secondary comparison.

RESULTS: In total, 291 patients were included. Primary analysis was done on patients where both physicians agreed upon the pain classification (n = 228). Compared to the physician's classification, PainDETECT-Dlv had a sensitivity of 80% and specificity of 55%, versus the Grading System it achieved 74 and 46%.

CONCLUSION: Despite its internal consistency and test-retest reliability the PainDETECT-Dlv is not an effective screening tool for a neuropathic pain component in a population of patients with chronic pain because of its moderate sensitivity and low specificity. Moreover, the indiscriminate use of the PainDETECT-Dlv as a surrogate for clinical assessment should be avoided in daily clinical practice as well as in (clinical-) research. Catch-22 situations in the validation of screening tools can be prevented by not pre-stratifying the patients on basis of the target outcome before inclusion in a validation study for screening instruments.

TRIAL REGISTRATION: The protocol was registered prospectively in the Dutch National Trial Register: NTR 3030 .

Original language	English
Pages (from-to)	91
Journal	BMC Neurology
Volume	18
Issue number	1
DOIs	https://doi.org/10.1186/s12883-018-1094-4
Publication status	Published - 29 Jun 2018

Keywords

Adult
Aged
Chronic Pain/diagnosis
Cross-Sectional Studies
Female
Humans
Male
Middle Aged
Neuralgia/diagnosis
Reproducibility of Results
Sensitivity and Specificity
Surveys and Questionnaires

Access to Document

https://doi.org/10.1186/s12883-018-1094-4

Cite this

@article{dffae5d036d2421dae6c58c8a97058ad,

title = "Avoiding Catch-22: validating the PainDETECT in a in a population of patients with chronic pain",

abstract = "BACKGROUND: Neuropathic pain is defined as pain caused by a lesion or disease of the somatosensory nervous system and is a major therapeutic challenge. Several screening tools have been developed to help physicians detect patients with neuropathic pain. These have typically been validated in populations pre-stratified for neuropathic pain, leading to a so called {"}Catch-22 situation:{"} {"}a problematic situation for which the only solution is denied by a circumstance inherent in the problem or by a rule{"}. The validity of screening tools needs to be proven in patients with pain who were not pre-stratified on basis of the target outcome: neuropathic pain or non-neuropathic pain. This study aims to assess the validity of the Dutch PainDETECT (PainDETECT-Dlv) in a large population of patients with chronic pain.METHODS: A cross-sectional multicentre design was used to assess PainDETECT-Dlv validity. Included where patients with low back pain radiating into the leg(s), patients with neck-shoulder-arm pain and patients with pain due to a suspected peripheral nerve damage. Patients' pain was classified as having a neuropathic pain component (yes/no) by two experienced physicians ({"}gold standard{"}). Physician opinion based on the Grading System was a secondary comparison.RESULTS: In total, 291 patients were included. Primary analysis was done on patients where both physicians agreed upon the pain classification (n = 228). Compared to the physician's classification, PainDETECT-Dlv had a sensitivity of 80% and specificity of 55%, versus the Grading System it achieved 74 and 46%.CONCLUSION: Despite its internal consistency and test-retest reliability the PainDETECT-Dlv is not an effective screening tool for a neuropathic pain component in a population of patients with chronic pain because of its moderate sensitivity and low specificity. Moreover, the indiscriminate use of the PainDETECT-Dlv as a surrogate for clinical assessment should be avoided in daily clinical practice as well as in (clinical-) research. Catch-22 situations in the validation of screening tools can be prevented by not pre-stratifying the patients on basis of the target outcome before inclusion in a validation study for screening instruments.TRIAL REGISTRATION: The protocol was registered prospectively in the Dutch National Trial Register: NTR 3030 .",

keywords = "Adult, Aged, Chronic Pain/diagnosis, Cross-Sectional Studies, Female, Humans, Male, Middle Aged, Neuralgia/diagnosis, Reproducibility of Results, Sensitivity and Specificity, Surveys and Questionnaires",

author = "Hans Timmerman and Wolff, {Andr{\'e} P} and Bronkhorst, {Ewald M} and Wilder-Smith, {Oliver H G} and Schenkels, {Marcel J} and {van Dasselaar}, {Nick T} and Huygen, {Frank J P M} and Steegers, {Monique A H} and Vissers, {Kris C P}",

year = "2018",

month = jun,

day = "29",

doi = "https://doi.org/10.1186/s12883-018-1094-4",

language = "English",

volume = "18",

pages = "91",

journal = "BMC Neurology",

issn = "1471-2377",

publisher = "BioMed Central",

number = "1",

}

TY - JOUR

T1 - Avoiding Catch-22

T2 - validating the PainDETECT in a in a population of patients with chronic pain

AU - Timmerman, Hans

AU - Wolff, André P

AU - Bronkhorst, Ewald M

AU - Wilder-Smith, Oliver H G

AU - Schenkels, Marcel J

AU - van Dasselaar, Nick T

AU - Huygen, Frank J P M

AU - Steegers, Monique A H

AU - Vissers, Kris C P

PY - 2018/6/29

Y1 - 2018/6/29

N2 - BACKGROUND: Neuropathic pain is defined as pain caused by a lesion or disease of the somatosensory nervous system and is a major therapeutic challenge. Several screening tools have been developed to help physicians detect patients with neuropathic pain. These have typically been validated in populations pre-stratified for neuropathic pain, leading to a so called "Catch-22 situation:" "a problematic situation for which the only solution is denied by a circumstance inherent in the problem or by a rule". The validity of screening tools needs to be proven in patients with pain who were not pre-stratified on basis of the target outcome: neuropathic pain or non-neuropathic pain. This study aims to assess the validity of the Dutch PainDETECT (PainDETECT-Dlv) in a large population of patients with chronic pain.METHODS: A cross-sectional multicentre design was used to assess PainDETECT-Dlv validity. Included where patients with low back pain radiating into the leg(s), patients with neck-shoulder-arm pain and patients with pain due to a suspected peripheral nerve damage. Patients' pain was classified as having a neuropathic pain component (yes/no) by two experienced physicians ("gold standard"). Physician opinion based on the Grading System was a secondary comparison.RESULTS: In total, 291 patients were included. Primary analysis was done on patients where both physicians agreed upon the pain classification (n = 228). Compared to the physician's classification, PainDETECT-Dlv had a sensitivity of 80% and specificity of 55%, versus the Grading System it achieved 74 and 46%.CONCLUSION: Despite its internal consistency and test-retest reliability the PainDETECT-Dlv is not an effective screening tool for a neuropathic pain component in a population of patients with chronic pain because of its moderate sensitivity and low specificity. Moreover, the indiscriminate use of the PainDETECT-Dlv as a surrogate for clinical assessment should be avoided in daily clinical practice as well as in (clinical-) research. Catch-22 situations in the validation of screening tools can be prevented by not pre-stratifying the patients on basis of the target outcome before inclusion in a validation study for screening instruments.TRIAL REGISTRATION: The protocol was registered prospectively in the Dutch National Trial Register: NTR 3030 .

AB - BACKGROUND: Neuropathic pain is defined as pain caused by a lesion or disease of the somatosensory nervous system and is a major therapeutic challenge. Several screening tools have been developed to help physicians detect patients with neuropathic pain. These have typically been validated in populations pre-stratified for neuropathic pain, leading to a so called "Catch-22 situation:" "a problematic situation for which the only solution is denied by a circumstance inherent in the problem or by a rule". The validity of screening tools needs to be proven in patients with pain who were not pre-stratified on basis of the target outcome: neuropathic pain or non-neuropathic pain. This study aims to assess the validity of the Dutch PainDETECT (PainDETECT-Dlv) in a large population of patients with chronic pain.METHODS: A cross-sectional multicentre design was used to assess PainDETECT-Dlv validity. Included where patients with low back pain radiating into the leg(s), patients with neck-shoulder-arm pain and patients with pain due to a suspected peripheral nerve damage. Patients' pain was classified as having a neuropathic pain component (yes/no) by two experienced physicians ("gold standard"). Physician opinion based on the Grading System was a secondary comparison.RESULTS: In total, 291 patients were included. Primary analysis was done on patients where both physicians agreed upon the pain classification (n = 228). Compared to the physician's classification, PainDETECT-Dlv had a sensitivity of 80% and specificity of 55%, versus the Grading System it achieved 74 and 46%.CONCLUSION: Despite its internal consistency and test-retest reliability the PainDETECT-Dlv is not an effective screening tool for a neuropathic pain component in a population of patients with chronic pain because of its moderate sensitivity and low specificity. Moreover, the indiscriminate use of the PainDETECT-Dlv as a surrogate for clinical assessment should be avoided in daily clinical practice as well as in (clinical-) research. Catch-22 situations in the validation of screening tools can be prevented by not pre-stratifying the patients on basis of the target outcome before inclusion in a validation study for screening instruments.TRIAL REGISTRATION: The protocol was registered prospectively in the Dutch National Trial Register: NTR 3030 .

KW - Adult

KW - Aged

KW - Chronic Pain/diagnosis

KW - Cross-Sectional Studies

KW - Female

KW - Humans

KW - Male

KW - Middle Aged

KW - Neuralgia/diagnosis

KW - Reproducibility of Results

KW - Sensitivity and Specificity

KW - Surveys and Questionnaires

U2 - https://doi.org/10.1186/s12883-018-1094-4

DO - https://doi.org/10.1186/s12883-018-1094-4

M3 - Article

C2 - 29958535

SN - 1471-2377

VL - 18

SP - 91

JO - BMC Neurology

JF - BMC Neurology

IS - 1

ER -