| 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. |
| 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.