M ETHODOLOGICAL DISCUSSION

The results from this thesis will add to the knowledgebase on hand function, activity, and participation in patients with PM and DM. To describe the methods and results in the

thesis, the ICF level-one classification was used. The use of a level-one classification gives a broad description about what ICF components and categories in the ICF might be impaired in patients with PM and DM. Further studies would be of interest to enable disease-specific core sets for patients with myositis.

A general limitation of the studies in Papers I-III is that the measures of grip force, grip strength, and HRQoL were compared to reference values from the literature. There is a limitation of using literature-derived control groups. To handle this, a gender- and age-matched value was collected from the literature for every patient.

In Paper I, the mean value for this control group was used in the one-sample t-test analysis. In Papers II and III, to address the deficiency of a control group, the analyses were more appropriately done by standardizing the observed values (mean values and standard deviations) and testing the null hypothesis (=there are no differences) by using the Wilcoxon Signed-rank test. Even though there was the limitation of using normative values from the literature, the comparable values along with patient-reported

assessments provided a direction of whether the patients had disabilities or not.

The E-link pinch meter has a thinner profile than the regular pinch gauge meter and could therefore not be compared to normative values. In the future, a study to derive normative values for the E-link pinch meter would give valuable information about the pinch grip strength.

The measures of activity and participation (MAP, DASH) used in this thesis enabled a description of what limitations and restrictions are present in patients with PM and DM (89, 90, 104). However, these measures did not evaluate if the activity is meaningful (occupation), whether the participants actually perform it, or if it negatively affects their lives (61, 63, 64). Only a few articles have assessed activity limitation in patients with PM and DM (2, 33, 39, 145). Therefore, the results from this thesis will add to the

knowledgebase of activity and participation in patients with PM and DM.

One previous study showed that the MAP covers some of the activities that are important to improve for patients with PM and DM (146). However, some areas are still missing in the MAP, such as sexual activity, sleep, and bicycling (146). In the future, it would be of interest to investigate which activities limit patients and which are of importance.

A patient-derived activity limitation measure would be preferred both in research and in clinical practice.

In this thesis, the SF-36 was chosen to evaluate HRQoL. As with the MAP and DASH measures, there might be areas of HRQoL that are important to the patient but are not covered by the SF-36. Therefore, it is necessary to perform qualitative studies to explore areas where the patients need support so that the health care providers can help the patients in the most appropriate way.

The WAI was able to detect reduced work ability in patients with PM and DM. The WAI has been used to predict the use of health care and rehabilitation and has also been shown to have associations to HRQoL (147), suggesting the use of WAI as a screening tool to evaluate self-rated work ability in patients with PM and DM in clinical practice.

In Paper II, the SweMyoNet Registry was used to collect data on grip force and HRQoL over time from the time of diagnosis. Missing data in registries based on data collected through clinical practise are a known general limitation of registries. Other explanations for the missing values may include the medical condition of the patient or potentially missed values from patients with the most severe or mild disability. To handle missing values and small sample sizes in the analysis, the Bonferroni after test was used to modify significance levels. The mixed linear model was used to enable an analysis over time despite the missing values (117). Still, we cannot exclude that the missing values may have affected the results in either a or negative manner. Furthermore, the lower number of observations after two years makes data from the later time points less reliable, particularly when subgrouping in women and men.

The aim of Paper III was to develop a hand exercise programme and to primarily test its feasibility in patients with PM and DM. Even though information about its effect on hand function and activity performance would have been beneficial for clinical practice guidelines, these results were not to be expected with this small convenience sample.

However the study did give directions on how to improve the programme design and what measures to include in a future study. A limitation in Paper III was that at the time when the hand exercise study was initiated there were no guidelines on which resistance putty to use based on hand grip strength which nowadays are available. This resulted in a more

subjective estimation of the resistance in the dough and for the participants a too time-consuming programme.

Paper IV included patients both working and on sick leave for more than two years, enabling a possible comprehensive picture about work ability in patients with PM and DM.

However, there might be a recall bias since the patients that had been on sick leave for more than 2 years had a range up to 162 months (≈14 years) of sick leave.