Part1: Introduction

There are 6 series of questions. Let us denote them by q1 to q6.

For activities done by right/left hand, we use “a”/“b” to denote it; for activities that require both hand, we use “c” to denote it.

Questions q31 to q35 are about how the patient’s work is influenced by the injury.

Part2: Pre-processing

Reoder

Recode the coding so that 1 means “condition is best”, 5 means “condition is worst”.

Rearrange by “dominant & injury”

We recode the data and question by

  • “a” means “dominant hand”
  • “b” means “non dominant hand”

Then the question change from “how is you right hand?” to “how is you dominant hand?”

Missingness

For Question q4a1: “How often did you have pain in your right hand(s)/wrist(s)?” , If the answer is “never”, then the following answer about “how pain influenece you” will miss. Thus I recode them as the “best condition” due to there being no pain. ( Is that OK? )

Part3: Partition into two groups

In the following, we divide the cases to two groups: when “Injury hand is dominant hand” and when “Injury hand is not dominant hand”.

For every people, we only consider injury hand related activities (questions about the patient’s injury hand or both hand).

All the following analysis is done using R package “eRm”.

First group: injury hand = dominant hand

Statistical result

## mhq1a2 mhq1a5 mhq2c1 mhq2c5 mhq5a1 mhq5a2 mhq5a3 mhq6a6 
##  0.029  0.038  0.002  0.002  0.000  0.000  0.000  0.094

According to hypothesis testing, these questions seems to be out of fit, i.e., they don’t reflect the true condition well.

Analysis

Let’s look into every question that doesn’t decribe well (green color):

  • q1a2: How well did your injury hand fingers move?

  • q2c1: Open a Jar

  • q2c5: Wash dishes

  • q1a5 : How was the sensation (feeling) in your injury hand?

  • q5a1: I am satisfied with the appearance(look) or my injury hand.

  • q5a2: The appearance (look) of my injury hand sometimes made me uncomfortable in public.

  • q5a3: The appearance (look) of my injury hand made me depressed.

  • q5a4: This one is OK: The appearance (look) of my injury hand interfered with my normal social activities .

  • q6a6 : Sensation (feeling) of your injury hand

For question q1a2, possible interpretation is that maybe finger movement is relatively independent of hand condition.

For question q2c1, possible interpretation is that the difficulty of this action also depend on sex, stength and other confounders. Or this task is too hard, and people have different critetria about relative “simple” or “hard”.

For question q2c5, maybe some have dish washer while others not?

For the other problematic questions, they share one common property: they all depends on whether the judgement is objective or not. Talking about “feeling” is in fact quite subjective. As comparison, problem q5a4 is relatively more subjective(it is about interference), thus still fits the model well while others not.

  • One personal question, what’s the difference between q1a5 and q6a6 ?

Second group: injury hand \(\not =\) dominant hand

Statistical result

## mhq2c1  mhq31 mhq5b1 mhq5b2 mhq5b3 mhq5b4 
##  0.000  0.001  0.000  0.000  0.000  0.000

Analysis

  • q2c1: Open a Jar

  • q31: How often were you unable to do your work because of problems with your hand(s)/wrist(s)?

  • q5b1 , q5b2, q5b3, q5b4 same question as (q5a1,q5a2,q5a3)

Interestingly, for those whose injury hand is not dominant hand, different from the other group, question q31 don’t reflect the true condition well.

It’s possible that “injury hand is not dominant hand” is a strong confounder: those associated with manual labor (which require both hand to work) are usually more likely to hurt their non-dominant hand, while teachers or managers take less risk.

Questions fitting condition

Questions with red color means out of fit, and questions with green color means fitting well.

Series 1 injury

  • 1. Overall, how well did your injury hand work?
  • 2. How well did your injury fingers move?
  • 3. How well did your injury wrist move?
  • 4. How was the strength in your injury hand?
  • 5. How was the sensation (feeling) in your injury hand?

Series 2, injury hand

  • 1. Turn a door knob
  • 2. Pick up a coin
  • 3. Hold a glass of water
  • 4. Turn a key in a lock
  • 5. Hold a frying pan

Series 2, both hand

  • 1. Open a Jar
  • 2. Button a shirt/blouse
  • 3. Eat with a knife/fork
  • 4. Carry a grocery bag
  • 5. Wash dishes
  • 6. Wash your hair
  • 7. Tie shoelaces/knots

Series 3

  • 1. How often were you unable to do your work because of problems with your hand(s)/wrist(s)?
  • 2. How often did you have to shorten your work day because of problems with your hand(s)/wrist(s)?
  • 3. How often did you have to take it easy at your work because of problems with your hand(s)/wrist(s)?
  • 4. How often did you accomplish less in your work because of problems with your hand(s)/wrist(s)?
  • 5. How often did you take longer to do the tasks in your work because of problems with your hand(s)/wrist(s)?

Series 4

  • 1. How often did you have pain in your injury hand(s)/wrist(s)?
  • 2. Please describe the pain you had in your injury hand(s)/wrist(s)
  • 3. How often did the pain in your injury hand(s)/wrist(s) interfere with your sleep?
  • 4. How often did the pain in your injury hand(s)/wrist(s) interfere with your daily activities (such as eating or bathing)?
  • 5. How often did the pain in your injury hand(s)/wrist(s) make you unhappy?

Series 5

  • 1. I am satisfied with the appearance(look) or my injury hand.
  • 2. The appearance (look) of my injury hand sometimes made me uncomfortable in public.
  • 3. The appearance (look) of my injury hand made me depressed.
  • 4. The appearance (look) of my injury hand interfered with my normal social activities.

Series 6

  • 1. Overall function of your injury hand
  • 2. Motion of the fingers in your injury hand
  • 3. Motion of your injury wrist
  • 4. Strength of your injury hand
  • 5. Pain level of your injury hand
  • 6. Sensation (feeling) of your injury hand

Part 4: Our questions about the data:

We decide to use “degree of Injury hand” - “degree of normal hand” as score to relieve the personal difference, and then fit the model. (Girls can hardly open a jar, regardless injury or not.)

But there seems to be some strange data. For example, For patients “OC24”, whose dominant hand (right hand) is injured, but his right hand (injured hand) self-reported result is “very good”, while left hand (normal hand ) is only “good”. There are many such conditions.