ABILHAND - Instructions

site navigation
  1. Downloading test packages
  2. Administering the questionnaire
  3. Analysing the subject's responses
  4. Interpreting the manual ability measures

^^^

Downloading test packages

The test packages for administrating the ABILHAND questionnaire are available for download to registered users. Registration is FREE and will only take a few minutes to complete. The test packages are available in various languages.

Each package contains the instruction sheet, the 10 scoring sheets (10 different random orders) and the response scale that is presented to the patient during the evaluation.

^^^

Administering the questionnaire

Figure 1: ABILHAND scoring sheet [+] Figure 1: ABILHAND scoring sheet

The ABILHAND questionnaire is administered on an interview basis.

The patient is asked to estimate the ease or difficulty of performing each activity when the activities are done without help, irrespective of the limb(s) the patient actually uses and whatever the strategies used to perform the activity. Note that the patient is never asked to perform the activities in front of the evaluator.

The activities are presented in a random order to avoid any systematic effect. 10 different random orders of presentation are used. The evaluator selects the next one of the 10 orders for each new assessment, no matter which patient is tested.

During the evaluation, the 3-level response scale is presented to the patient. The patient is asked to rate his/her perception on the response scale as "Impossible", "Difficult" or "Easy". The activities not attempted within the last 3 months are not scored and are entered as not applicable (check the question mark "?" on the scoring sheet). The activities that the patient does not perform because they are too difficult must be scored as "Impossible".

The scores are entered as shown in the Figure 1.

^^^

Analysing the subject's responses

The responses to the questionnaire can be submitted for an online analysis on this web site.

The analysis uses the Rasch model to convert the raw scores into a linear measure. The linear measure can be determined even when some items have not been answered (missing responses will only reduce the precision of the measure).

The online form is available with the items ordered in the 10 random orders used on the paper scoring sheets. Select the language and the order (1 to 10) that were used during the evaluation. Then enter the patient's responses into the online form and submit the form to display the evaluation report.

^^^

Interpreting the manual ability measures

Figure 2: ABILHAND evaluation report [+] Figure 2: ABILHAND evaluation report

The evaluation report is presented in Figure 2.

The patient's manual ability (vertical red line) and its 95% confidence interval (dotted red lines) are located on the manual ability scale (abscissa), expressed in logits. The higher the ability of a patient the more the measure will be located to the right.

The logit is a linear unit that expresses the odds of success of the patient on any given item. The manual ability scale is centred on the average item difficulty (0 logit).

The item map lists the items from top to bottom in decreasing difficulty order. The item map shows the most probable score (0 = "Impossible", 1 = "Difficult" or 2 = "Easy") to each item as a function of the patient's manual ability and the difficulty of the item. Note that "Hammering a nail" is the most difficult item as it requires the highest ability to be succeeded easily; "Washing one's hands" is the easiest. Note also that the most probable score for any given item increases with the patient's manual ability. The item map can be used for treatment planning and follow-up as it shows the item(s) that should be recovered first during the rehabilitation process (obviously, the easier ones).

The sigmoid curve shown at the bottom relates the total score on the questionnaire to the manual ability measure (for a complete response set). Note that the score always increases with the manual ability, although, given the shape of the curve, the increase is not linear. Therefore changes of manual ability must be interpreted in linear logit units rather than in raw scores.

The patient's response to each item is circled (missing responses are not shown). In this particular case, the patient can still recover performance in most of the difficult items down to "Pulling up the zipper of trousers". The patient should first recover easier activities like "Pulling up the zipper of trousers" up to more difficult activities like "Hammering a nail".

For any particular item, the range overlapping the 95% confidence interval indicates the most probable score(s) given the overall manual ability measure. Note that in the example, most scores fit with the overall manual ability measure of the patient indicating the patient scores are coherent across items. The item "Buttoning up trousers" is scored as "Impossible" while it is expected to be easier, given the overall manual ability measure. The item "Tearing open a pack of chips" is scored as "Easy" while it is expected to be more difficult, given the overall manual ability measure. These unexpected scores (circled in red on the report) may help you diagnose an atypical behaviour of the patient or other comorbidities.

The evaluation report also includes the numeric values for the patient's manual ability measure on the scale and the standard error of measurement (in logits). Keep those in your records to measure the patient recovery at a future occasion.

^^^