Limits...
Validation of an informant-reported web-based data collection to assess dementia symptoms.

Rockwood K, Zeng A, Leibman C, Mucha L, Mitnitski A - J. Med. Internet Res. (2012)

Bottom Line: In the 250 profilees (155, 62% women, mean age 77 years), increasing dependence was associated with a greater chance of institutionalization.For example, no one at the lowest levels of dependence (DS score < 5, n = 33) was in long-term care, compared with half (13/25) of the profilees at the highest levels of dependence (DS score > 12) being in institutions (χ(2)(4) = 27.9, P < .001).Even so, caution is needed to assure the validity of data collected online.

View Article: PubMed Central - HTML - PubMed

Affiliation: DGI Clinical Inc, Halifax, NS, Canada. KRockwood@dgiclinical.com

ABSTRACT

Background: The Web offers unprecedented access to the experience of people with dementia and their care partners, but data gathered online need to be validated to be useful.

Objective: To test the construct validity of an informant Web-based data collection to assess dementia symptoms in relation to the 15-point Dependence Scale (DS).

Methods: In an online survey posted on the DementiaGuide website, care partners of people with dementia built individualized profiles from the 60-item SymptomGuide and completed a questionnaire, which included the DS and a staging tool.

Results: In the 250 profilees (155, 62% women, mean age 77 years), increasing dependence was associated with a greater chance of institutionalization. For example, no one at the lowest levels of dependence (DS score < 5, n = 33) was in long-term care, compared with half (13/25) of the profilees at the highest levels of dependence (DS score > 12) being in institutions (χ(2)(4) = 27.9, P < .001). The Web-based DS was correlated with the number of symptoms: higher DS scores were associated with a higher stage of dementia (F > 50, P < .001).

Conclusion: In an online survey, the Web-based DS showed good construct validity, potentially demonstrating how the Web can be used to learn more about dementia progression and how it relates to symptoms experienced by patients across the course of dementing illnesses. Even so, caution is needed to assure the validity of data collected online.

Show MeSH

Related in: MedlinePlus

Probability distribution of the Dependence Scale score by dementia stage. Arrows indicate crossovers between the neighboring distributions and thus represent the the cut points to distinguish between the stages.
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC3376520&req=5

figure4: Probability distribution of the Dependence Scale score by dementia stage. Arrows indicate crossovers between the neighboring distributions and thus represent the the cut points to distinguish between the stages.

Mentions: We discerned 5 empirically derived cut points at ≤5, ≤8, ≤11, ≤13, and >13; the modes increased with increasing values of the DS (Figure 4) as did the median values (ie, ≤5, median proportion value of the DS; ≤8; ≤11; ≤13; >13). We chose the cut points by using the crossovers between the neighboring distributions as indicated in the legend of Figure 4. The distributions are not normal (at least on the extremes, the medians are deviate from the means); the crossover’s location, however, indicates that on the right from the crossover, the probability of the right distribution is higher than that on the left. It could be said that the cut points were chosen in accordance with the maximum likelihood of belonging to one or the other group. Note that when the DS score was <2, profilees were almost all in the stage of very mild dementia. When the DS score was between 2 and 8, most profilees were in the mild dementia stage. Profilees identified as having moderate dementia had a DS score varying from 6 to 12. When the DS score was >10, the profilees were predominantly at the severe or very severe stage of dementia, and no one remained in the mild stage.


Validation of an informant-reported web-based data collection to assess dementia symptoms.

Rockwood K, Zeng A, Leibman C, Mucha L, Mitnitski A - J. Med. Internet Res. (2012)

Probability distribution of the Dependence Scale score by dementia stage. Arrows indicate crossovers between the neighboring distributions and thus represent the the cut points to distinguish between the stages.
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC3376520&req=5

figure4: Probability distribution of the Dependence Scale score by dementia stage. Arrows indicate crossovers between the neighboring distributions and thus represent the the cut points to distinguish between the stages.
Mentions: We discerned 5 empirically derived cut points at ≤5, ≤8, ≤11, ≤13, and >13; the modes increased with increasing values of the DS (Figure 4) as did the median values (ie, ≤5, median proportion value of the DS; ≤8; ≤11; ≤13; >13). We chose the cut points by using the crossovers between the neighboring distributions as indicated in the legend of Figure 4. The distributions are not normal (at least on the extremes, the medians are deviate from the means); the crossover’s location, however, indicates that on the right from the crossover, the probability of the right distribution is higher than that on the left. It could be said that the cut points were chosen in accordance with the maximum likelihood of belonging to one or the other group. Note that when the DS score was <2, profilees were almost all in the stage of very mild dementia. When the DS score was between 2 and 8, most profilees were in the mild dementia stage. Profilees identified as having moderate dementia had a DS score varying from 6 to 12. When the DS score was >10, the profilees were predominantly at the severe or very severe stage of dementia, and no one remained in the mild stage.

Bottom Line: In the 250 profilees (155, 62% women, mean age 77 years), increasing dependence was associated with a greater chance of institutionalization.For example, no one at the lowest levels of dependence (DS score < 5, n = 33) was in long-term care, compared with half (13/25) of the profilees at the highest levels of dependence (DS score > 12) being in institutions (χ(2)(4) = 27.9, P < .001).Even so, caution is needed to assure the validity of data collected online.

View Article: PubMed Central - HTML - PubMed

Affiliation: DGI Clinical Inc, Halifax, NS, Canada. KRockwood@dgiclinical.com

ABSTRACT

Background: The Web offers unprecedented access to the experience of people with dementia and their care partners, but data gathered online need to be validated to be useful.

Objective: To test the construct validity of an informant Web-based data collection to assess dementia symptoms in relation to the 15-point Dependence Scale (DS).

Methods: In an online survey posted on the DementiaGuide website, care partners of people with dementia built individualized profiles from the 60-item SymptomGuide and completed a questionnaire, which included the DS and a staging tool.

Results: In the 250 profilees (155, 62% women, mean age 77 years), increasing dependence was associated with a greater chance of institutionalization. For example, no one at the lowest levels of dependence (DS score < 5, n = 33) was in long-term care, compared with half (13/25) of the profilees at the highest levels of dependence (DS score > 12) being in institutions (χ(2)(4) = 27.9, P < .001). The Web-based DS was correlated with the number of symptoms: higher DS scores were associated with a higher stage of dementia (F > 50, P < .001).

Conclusion: In an online survey, the Web-based DS showed good construct validity, potentially demonstrating how the Web can be used to learn more about dementia progression and how it relates to symptoms experienced by patients across the course of dementing illnesses. Even so, caution is needed to assure the validity of data collected online.

Show MeSH
Related in: MedlinePlus