The Canadian minimum dataset for chronic low back pain research: a cross-cultural adaptation of the National Institutes of Health Task Force Research Standards

Anaïs Lacasse; Jean-Sébastien Roy; Alexandre J. Parent; Nioushah Noushi; Chúk Odenigbo; Gabrielle Pagé; Nicolas Beaudet; Manon Choinière; Laura S. Stone; Mark A. Ware; for the Quebec Pain Research Network's Steering Committee of the Low Back Pain Strategic Initiative

doi:10.9778/cmajo.20160117

Article Figures & Tables

Tables

View popup

Table 1: Overview of the adaptations made to the original minimum data set and endorsed by the expert committee

Items of the original minimum data set		French translation method	Adaptations made to both French and English versions
Human body drawing to locate low back pain		NA	Addition: Introduction sentence to inform the respondent that the drawing shows the part of the body referred to in the questionnaire
1,2	LBP duration	Double forward-backward translation	Modification (item 1): Answer choices "1-3 months", "3-6 months" and "6 months-1 year" were respectively replaced by "1-2 months", "3-5 months" and "6-11 months" to allow a collectively exhaustive set of answers
3	LBP intensity in the past 7 days	Double forward-backward translation	A typographical error was made by the authors of the minimum data set when presenting the pain intensity scale (1-10 rather than 0-10 numerical rating scale);^14-22 the correction was made because 0-10 scales are considered a standard in pain research²⁹
4	Sciatica	Double forward-backward translation	Addition: Human body drawing to reduce uncertainty and missing data Standardization of the item: "low-back pain" instead of "back pain"
5	Comorbid pain conditions	Double forward-backward translation	NA
6, 7, 8	History of LBP surgical interventions	Double forward-backward translation	Modification (items 6-8): "surgery" instead of "operation" Modification (item 7): The second answer choice "More than 6 months, but less than 1 year ago" was replaced by "6 months or more but less than 1 year ago" to allow a collectively exhaustive set of answers Standardization of items: "low-back surgery" instead of "back surgery"
9, 10, 11, 12	Pain interference	French version ordered directly from PROMIS Cooperative Group²⁴	NA
13	LBP treatments	Double forward-backward translation	Modification: The list of examples of prescription medications was adjusted according to drugs available in Canada Standardization of items: "low-back pain" instead of "back pain"
14	LBP-related workplace absenteeism	Double forward-backward translation	Modification: The answer choices "Agree" and "Disagree" were replaced by "Yes" and "No"
15	LBP-related workers' compensation benefits	Double forward-backward translation	Modification: The answer choices "Agree" and "Disagree" were replaced by "Yes" and "No"
16, 17, 18, 19	Physical function	French version ordered directly from PROMIS Cooperative Group²⁴ Addition: We felt that the French version of item 19 ("Are you able to run errands and shop?") used European French wording that is not straightforward for French-Canadians of various socioeconomic statuses; to reduce the risk of losing psychometric validity, the item was not reformulated but a bracketed specification was added to increase clarity	NA
20, 21, 22, 23	Emotional distress or depression	French version ordered directly from PROMIS Cooperative Group²⁴	NA
24, 25, 26, 27	Sleep disturbance	French version ordered directly from PROMIS Cooperative Group²⁴	NA
28	Kinesiophobia	French-European version of the STarT Back Screening Tool²⁵	Standardization of the item: "low-back problem" instead of "back problem"
29	Catastrophizing	French-European version of the STarT Back Screening Tool²⁵	Standardization of the item: "low-back pain" instead of "back pain"
30	LBP-related lawsuits and legal claims	Double forward-backward translation	Standardization of the item: "low-back problem" instead of "back problem"
31, 32	Substance abuse	Double forward-backward translation	Modification (item 31): We felt that the formulation lacked clarity about the type of substances the questionnaire refers to ("Have you drunk or used drugs more than you meant to?"); the sentence was replaced by "Have you consumed alcohol or used drugs more than you meant to?"
33	Age	Double forward-backward translation	NA
34	Sex	Double forward-backward translation	NA
35, 36	Race and ethnicity	Double forward-backward translation	Modification: Answer choices were replaced by non-mutually exclusive categories used in Statistics Canada Censuses (National Household Survey)³⁰ that are available in French and English
37	Employment Status	Double forward-backward translation	NA
38	Highest education level attained	Double forward-backward translation	Modification: Answer choices were replaced by mutually exclusive categories used in Statistics Canada Censuses (National Household Survey)³⁰ that are available in French and English; additional examples were added by our team for some education-level categories
39	Smoking status	Double forward-backward translation	NA
40	Obesity (height and weight)	Double forward-backward translation	The presentation format of the questions was slightly modified to ease data collection and analysis; the specifications regarding whether height and weight were measured or self-reported were replaced by "has just been measured" and "is an estimation", respectively

Note: LBP = low back pain, NA = not applicable, PROMIS = Patient-Reported Outcomes Measurement Information System.

View popup

Table 2: Characteristics of the study participants (n = 33)

Characteristic	No. (%)*
Age, yr, mean ± SD†	46.5 ± 14.4
Median	41
Minimum	24
Maximum	82
Sex‡
Female	21 (63.6)
Male	12 (36.4)
Unknown or unspecified	0 (0.0)
Aboriginal persons§ (not mutually exclusive categories)
Not an Aboriginal person	30 (96.8)
First Nations (North American Indian)	1 (3.2)
Other categories listed (Métis or Inuk)	0 (0.0)
Race/ethnicity¶ (not mutually exclusive categories)
White	31 (96.9)
Arab	2 (6.3)
Other categories listed	0 (0.0)
Employment status‡ (not mutually exclusive categories)
Working now (full- or part-time)	22 (66.7)
Retired	7 (21.2)
Disabled due to back pain (permanently or temporarily)	3 (9.1)
Student	1 (3.0)
Looking for work	1 (3.0)
Other categories listed	0 (0.0)
Highest education level attained‡
No high school diploma	3 (9.1)
High school diploma or equivalent	4 (12.1)
Registered apprenticeship or other trades certificate or diploma	3 (9.1)
College, CEGEP or other non-university certificate or diploma	6 (18.2)
University certificate or diploma below bachelor's level	0 (0.0)
Bachelor's degree	7 (21.2)
University certificate or diploma above bachelor's level	3 (9.1)
Master's degree	4 (12.1)
Degree in medicine, dentistry, veterinary medicine or optometry	0 (0.0)
Doctorate	3 (9.1)
LBP that persisted at least 3 mo‡
Yes	30 (90.9)
No	3 (9.1)
LBP that has resulted in pain on at least half the days in the past 6 mo§
Yes	16 (51.6)
No	15 (48.4)
Presence of CLBP based on the NIH definition§
Yes	15 (48.4)
No	16 (51.6)
LBP intensity in the past 7 d (0-10), mean ± SD‡	4.8 ± 2.3
Median	5 ± 2.3
Minimum	0
Maximum	9

Note: CLBP = chronic low back pain, LBP = low back pain, SD = standard deviation.

*Unless otherwise specified.†5 missing data.

‡0 missing data.

§2 missing data.

¶ 1 missing data.

View popup

Table 3: Overview of results of the pretest, and changes made to produce the final version of the Canadian minimum data set

Items of the original minimum data set		Proportion of participants reporting the item as completely clear (n = 35), %	Comments and suggestions of participants who reported that the clarity of the item should be improved	Modifications brought to the final version of the French and English questionnaires
Human body drawing to locate low back pain		100.0	NA	NA
1, 2	LBP duration	91.4	Participants reported that they have LBP, but that it is not a problem for them. They suggested clarifying the term "problem" (n = 3).	Considering the small proportion of participants who reported that these items were unclear and the importance of these 2 items for the NIH's standard definition of CLBP cases, no modification was made to these 2 items.
1, 2	LBP duration	94.3	They did not understand the relevance of the second question when the first one was already asked (n = 1). The 2 last answer choices of the second item are very similar (n = 1).
3	LBP intensity in the past 7 days	94.3	It is difficult to report LBP intensity in the past 7 days on average (n = 1). There are too many points on the scale (n = 1).	Because 0-10 numerical rating scales for average pain intensity are standards in pain research,²⁹ no modification was made.
4	Sciatica	94.3	The question should present a larger time window because 2 weeks is not representative of their condition (n = 2).	Given that comments did not reflect a clarity issue and in order to maximize the comparability of results obtained with the Canadian and American version of the minimum data set, the time frame was not modified.
5	Comorbid pain conditions	91.4	The difference between "widespread pain" and "pain in most of your body" as 2 different concepts was not clear (n = 1). It was not clear if pain in legs and arms could qualify as "pain in most of your body" (n = 1). Other pain conditions were not listed (n = 1).	To reduce confusion, "pain in most of your body" was presented in brackets as a synonym of "widespread pain" rather than a different concept.
6, 7, 8	History of LBP surgical interventions	77.1	The first question should be used as a branching question (screener) for items #7 and #8 that are follow-up questions for patients who answered "yes" to item #6. No instructions are provided (e.g., If no, go to question #9). They do not want to read questions that do not apply to them (n = 8). Just like for item #7, add "If yes" to item 8 (n = 1).	Future users of the minimum data set should number the items and use a skip ahead instruction (paper-and-pencil questionnaire) or use an automatic Question Skip Logic (web-based questionnaire).
9, 10, 11, 12	Pain interference	77.1	The difference between item 10 about "work around the home" and item 12 about "household chores" is unclear. Examples should be provided (n = 8). Gardening can be considered a day-to-day activity (item 9), work around the home (item 10) and a household chore (item 12). Where should we include this activity? (n = 1).	The research team was confronted with many comments regarding the PROMIS Pain interference scale but had to keep in mind that items of a validated scale cannot be reformulated without risk of losing psychometric validity. We however felt that giving examples to patients could be a good idea. Examples such as garden work and renovations were added for the "work around the home" item. Examples such as house cleaning and vacuuming were added for the "household chores" item.
13	LBP treatments	77.1	A time window should be presented, e.g., current use of treatments, use of treatments in the past year (n = 3). The general format of the question was not clear, i.e., same answer choices for the type of treatment used by the patient and for the current use of opioid painkillers (n = 3). The French term "Infiltration" should be added as a synonym of injections (n = 1). "Exercise therapy" is not clear. Does this include personal exercise plan that was not prescribed by a physiotherapist? (n = 1). It is not clear if injections and provided examples include intramuscular or intravenous methotrexate injections (n = 1).	Answer choices for treatment use were modified as follows: "Yes, I am currently using this treatment", "Yes, I have used this treatment in the past but stopped", "No", and "Not sure". The term "Infiltration" was added as a synonym of "Injections".
14	LBP-related workplace absenteeism	94.3	A time window should be presented (n = 2).	No modification was made. However, if a distinction between past and recent disability is important, researchers might add a time frame to this item (e.g., past 12 mo).
15	LBP-related workers' compensation benefits	100	NA	NA
16, 17, 18, 19	Physical function	94.3	A time window should be presented (n = 1). It is difficult to answer these questions because physical functioning varies according to pain symptoms during a given week (n = 1).	Considering the large proportion of participants who reported that these items were clear, and the importance of keeping the psychometric validity of the PROMIS Physical Function scale, no modification was made.
20, 21, 22, 23	Emotional distress/depression	94.3	The French term for "helpless" ("désemparé") is not clear (n = 1). The difference between the French term for "helpless" ("désemparé") and the French term for hopeless ("désespéré") is not clear (n = 1).	Aforementioned rationale. No modification was made.
24, 25, 26, 27	Sleep disturbance	94.3	It was not clear if the questions were about general or LBP-related sleep quality (n = 1). It was not clear if questions were about sleep quality with or without sleeping pills (n = 1).	Aforementioned rationale. No modification was made.
28	Kinesiophobia	85.7	Questions should not be formulated in a negative way (n = 3). The item is not clear. Do you mean that doing exercise will result in my condition getting worse? (n = 1). It was unclear according to whom physical activity was considered unsafe. The participant? His physician? Scientific evidence? (n = 1).	This item was taken from the French-European version of the STarT Back Screening Tool.²⁵ No modification was made.
29	Catastrophizing	94.3	There are 2 questions in this statement. It should be divided: LBP is terrible (agree/disagree) and LBP is never going to get any better (agree/disagree) (n = 1). This item should have a "not sure option" (n = 1).	This item was taken from the French-European version of the STarT Back Screening Tool.²⁵ No modification was made.
30	LBP-related lawsuits and legal claims	97.1	Examples of agencies where a legal claim can be submitted should be provided (n = 1).	Given the large proportion of participants who reported that this item was clear, and the variability of agencies across Canadian provinces, no modification was made.
31,32	Substance abuse	97.1	It was not clear if "drugs" would include only illicit drugs or also medications. Specifically mentioned about the debate surrounding opioid use (n = 1).	Given the large proportion of participants who reported that this item was clear, no modification was made.
33	Age	88.6	The specification "(0-120 yr)" sound weird/not required (n = 4).	No modification was made because this aspect should not affect the validity of the answer.
34	Gender	82.9	Added a question/exclamation mark at the side of the "Unknown" and "Unspecified" answer choices (n = 2). Both choices ("Unknown" and "Unspecified" answers) are not required. A simple "Other" category should be added (n = 4).	No modification was made following the pre-test in order to replicate the answer choices of the original data set.
35	Race and ethnicity	97.1	It was not clear if descendant aboriginal counts as an aboriginal person (n = 1).	Because it represented categories used in Statistics Canada Censuses,³⁰ no modification was made.
36		91.4	It was not clear if "White" meant "Quebecker" (n = 1). The "Arab" category should have examples. Is Maghreb included (n = 1)? Strange categories (n = 2).	Because it represented categories used in Statistics Canada Censuses,³⁰ no modification was made.
37	Employment status	85.7	The "working now" answer choice should be separated for full- and part-time work (n = 2). Added a question mark at the side of the "Unknown" answer choice (n = 1). Categories should not be mutually exclusive (n = 2).	The statement "mark more than one answer if applicable" was added, and full- versus part-time work are now presented as different options.
38	Highest education level attained	100	NA	NA
39	Smoking status	100	NA	NA
40	Obesity (height and weight)	85.7	The specification on whether height and weight have just been measured is not clear. When? By whom? (n = 5). It is strange and not relevant to ask if height and weight have just been measured or are estimations (n = 2).	The research team felt that the information about whether height and weight have been measured by the patient, by his physician or is a self-reported estimation was not of great relevance for data analysis and phenotyping of CLBP patients. It was thus removed. However, future users of the minimum data set could keep these specifications if they want to consider this information in their analysis.

Note: CLBP = chronic low back pain, LBP = low back pain, NA = not applicable, NIH = National Institutes of Health, PROMIS = Patient-Reported Outcomes Measurement Information System.