Measurement in Medicine: A Practical Guide (Practical Guides to Biostatistics and Epidemiology)

Title......Page 5
Copyright......Page 6
Contents......Page 7
Preface......Page 11
1.1 Why this textbook on measurement in medicine?......Page 13
1.3 Terminology and definitions......Page 14
1.4 Scope of measurements in medicine......Page 15
1.5 For whom is this book written?......Page 16
1.6 Structure of the book......Page 17
1.7 Examples, data sets, software and assignments......Page 18
2.2 Conceptual models......Page 19
From clinician-based to patient-based measurements......Page 22
From objective to subjective measurements......Page 23
From observable to non-observable characteristics......Page 24
2.4 Conceptual framework: reflective and formative models......Page 25
2.5 Measurement theories......Page 29
2.5.1 Classical test theory......Page 31
2.5.2 Item response theory......Page 32
2.6 Summary......Page 38
2. What is the construct?......Page 39
3. Item response theory......Page 41
3.1 Introduction......Page 42
3.2.1 Construct......Page 45
3.2.3 Purpose of measurement......Page 46
3.3 Choice of measurement method......Page 47
3.4.1.1 Literature......Page 49
3.4.1.2 Experts......Page 50
3.4.1.4 An example of item selection for a non-patient-reported outcomes instrument......Page 51
3.6.2 Summarizing scores in formative models......Page 63
3.4.2 Formulating items: first draft......Page 53
3.4.3 Things to keep in mind......Page 54
3.5.1 Scoring options......Page 58
3.5.2 Which option to choose?......Page 60
3.6 Scores for scales and indexes......Page 61
3.6.1 Summarizing scores in reflective models......Page 62
3.6.3.1 How and by who are weights assigned......Page 65
Empirical weights......Page 66
3.6.3.2 Preference weighting or utility analysis......Page 68
3.7 Pilot-testing......Page 69
3.7.1 Pilot-testing of patient-reported outcomes instruments......Page 70
3.8 Summary......Page 72
1. Definition of a construct......Page 73
3. Choice between a reflective and a formative model......Page 74
5. Use of sum-scores......Page 75
4.1 Introduction......Page 77
4.2.1 Missing scores......Page 78
4.2.2 Distribution of item scores......Page 80
4.3 Importance of the items......Page 82
4.4 Examining the dimensionality of the data: factor analysis......Page 83
4.4.1 Principles of exploratory factor analysis......Page 84
4.4.2 Determining the number of factors......Page 86
4.4.2.2 Step 2: the number of factors to be extracted......Page 87
4.4.4 Optimizing the dimensionality......Page 89
4.4.5 Some remarks on factor analysis......Page 91
4.5 Internal consistency......Page 92
4.5.2 Cronbach’s alpha......Page 93
4.5.3 Interpretation of Cronbach’s alpha......Page 95
4.6 Examining the items in a scale with item response theory......Page 96
4.6.1 Fit of an item response theory model......Page 97
4.6.2 Distribution of items over the scale......Page 99
4.7 Field-testing as part of a clinical study......Page 103
4.8 Summary......Page 104
1. Methods of item selection......Page 105
3. Factor analyses of the Graves’ ophthalmopathy quality of life questionnaire......Page 106
4. Cronbach’s alpha: Graves’ ophthalmopathy quality of life questionnaire......Page 107
5.1 Introduction......Page 108
5.3 The concept of reliability......Page 110
5.4.1.1 Intraclass correlation coefficients for single measurements......Page 115
5.4.1.2 Intraclass correlation coefficients for averaged measurements......Page 119
5.4.1.3 Pearson’s r......Page 122
5.4.2.1 Standard error of measurement......Page 123
5.4.2.2 Limits of agreement (Bland and Altman method)......Page 125
5.5.1.1 Cohen’s kappa for nominal variables......Page 127
Cohen’s kappa......Page 128
Weighted kappa......Page 129
5.5.2 No parameters of measurement error for categorical variables......Page 131
5.6.1.2 Kappa......Page 132
5.6.2.2 Bland and Altman method......Page 134
Question 1......Page 144
Question 2......Page 145
Question 3......Page 146
5.7 Which parameter to use in which situation?......Page 135
5.8 Design of simple reliability studies......Page 136
5.9 Sample size for reliability studies......Page 138
5.10 Design of reliability studies for more complex situations......Page 140
5.11.1 Generalizability studies......Page 143
Question 4......Page 147
5.12 Cronbach’s alpha as a reliability parameter......Page 149
5.13 Reliability parameters and measurement error obtained by item response theory analysis......Page 151
5.14 Reliability and computer adaptive testing......Page 153
5.15 Reliability at group level and individual level......Page 154
5.16 Improving the reliability of measurements......Page 156
5.17 Summary......Page 157
1. Calculation and interpretation of intraclass correlation coefficient......Page 158
4. Calculation of kappa......Page 159
6. Design of a generalizability study......Page 160
7. Exercise on generalizability and decision studies......Page 161
6.1 Introduction......Page 162
Knowledge about the construct......Page 163
Dependency on the situation......Page 164
Validation as a continuous process......Page 165
6.3 Content validity (including face validity)......Page 166
6.3.2 Content validity......Page 167
2: Consider information about content of the measurement instrument......Page 168
4: Assess whether the content of the measurement instrument corresponds with the construct......Page 169
5: Use a strategy or framework to assess the correspondence between the instrument and construct......Page 170
6.4 Criterion validity......Page 171
1: Identify a suitable criterion and method of measurement......Page 172
3: Define a priori the required level of agreement between measurement instrument and criterion......Page 173
5: Determine the strength of the relationship between the instrument scores and criterion scores......Page 174
Example of concurrent validity (dichotomous outcome)......Page 176
Example of concurrent validity (continuous outcome)......Page 178
6.4.2 Predictive validity......Page 179
6.5.1 Structural validity......Page 181
Example of confirmatory factor analysis......Page 182
6.5.2 Hypotheses testing......Page 184
2: Formulate hypotheses......Page 185
4: Gather empirical data......Page 186
Example of discriminative validation......Page 187
Example of convergent, discriminant and discriminative validation......Page 188
Example of hypotheses testing of the COOP/WONCA scales......Page 191
6.5.3 Cross-cultural validity......Page 193
Step 1: Forward translation......Page 194
Step 3: Back translation......Page 195
6.5.3.2 Cross-cultural validation......Page 196
Factor analysis......Page 197
Logistic regression analysis......Page 198
Example......Page 199
Item response theory techniques......Page 200
Example......Page 201
6.6.1 Sample sizes......Page 203
6.6.3 Validation along clinical studies......Page 204
6.6.4 Link with new clinical knowledge......Page 205
6.6.6 Validity and reliability......Page 206
6.7 Summary......Page 208
2. Drawing conclusions about construct validity......Page 209
4. Validation of a Short-Form version of the WOMAC versus the Long-Form version......Page 210
5. Interpretation of data on measurement invariance (differential item functioning)......Page 211
7.1 Introduction......Page 214
7.2.2 Definition of responsiveness......Page 215
7.2.3 Implications for measuring responsiveness......Page 216
7.3 Criterion approach......Page 218
7.3.1 Example of a continuous variable as gold standard......Page 220
7.3.2 Example of dichotomous variable as gold standard......Page 221
7.4 Construct approach......Page 223
Example of hypotheses testing......Page 224
7.5.1 Effect sizes......Page 227
7.5.2 Paired t-test......Page 228
7.5.3 Guyatt’s responsiveness ratio......Page 229
7.5.4 Exceptions......Page 230
7.6 Other design issues......Page 232
7.7 Summary......Page 233
Methods to assess responsiveness......Page 235
8.1 Introduction......Page 239
8.3.1 Importance of examining the distribution of scores......Page 240
8.3.2 Examining the distribution of scores using classical test theory methods......Page 242
8.3.3 Examining the distribution of scores using item response theory methods......Page 243
8.3.4 Floor and ceiling effects......Page 244
8.4 Interpretation of single scores......Page 247
8.4.1 Using the norm scores of a general population......Page 248
8.4.3 Interpretations of the scores of item response theory-based instruments......Page 249
8.4.4 Criteria for what is considered normal......Page 251
8.4.4.3.  Based on what is treatable......Page 252
8.5.2 Relationship with change scores on other known instruments......Page 253
8.5.3.1 Smallest detectable change is based on measurement error......Page 254
8.5.3.2 Smallest detectable change in individual patients and in groups of patients......Page 256
8.5.4.2 Methods to determine minimal important change......Page 257
Step 3: Determine the cut-off point......Page 260
An example......Page 261
Step 2: Plot the distribution of change scores......Page 262
Step 3: Determine the cut-off point......Page 264
Interpreting the anchor-based minimal important change distribution......Page 265
Minimal important change depends on the type of anchor......Page 268
Minimal important change depends on baseline values and patient groups......Page 269
8.5.5 Distinction between the smallest detectable change and the minimal important change......Page 270
8.5.6 Response shift......Page 273
8.5.6.1 A conceptual model of response shift......Page 275
Qualitative methods......Page 276
Quantitative methods......Page 277
Factor analysis......Page 278
8.5.6.3 Interpretation of response shift......Page 279
8.6 Summary......Page 280
2. Determining the minimal important change for the PRAFAB questionnaire, using the anchor-based minimal important change dist......Page 282
3. Response shift......Page 284
9.1 Introduction......Page 287
9.2.1 Types of systematic reviews......Page 288
9.2.3 Systematic reviews to select the best measurement instrument......Page 289
9.3 Literature search......Page 290
1: Construct of interest......Page 291
3: Measurement properties......Page 292
Example: Mazaheri et al. (2010)......Page 298
9.3.4 Publication bias......Page 293
9.4 Eligibility criteria......Page 294
9.5 Selection of articles......Page 295
9.6 Evaluation of the methodological quality of the included studies......Page 296
9.7.1 General characteristics of the instrument......Page 303
9.7.2 Characteristics of the study population......Page 304
9.8 Content comparison......Page 306
9.9.1 Homogeneity of the study characteristics......Page 308
9.9.3 Qualitative analysis (best evidence synthesis)......Page 309
Reliability......Page 310
Construct validity and responsiveness (hypotheses testing)......Page 311
2: The methodological quality of each study......Page 314
3: Characteristics of the included measurement instruments......Page 315
4. Characteristics of the study population......Page 317
6: Conclusion about the best measurement instrument......Page 318
9.10 Overall conclusions of the systematic review......Page 312
9.12 State of affairs......Page 321
9.13 Comprehensiveness of systematic reviews of measurement properties......Page 322
9.14 Summary......Page 323
2. Data synthesis of eight studies on the reliability of one instrument......Page 325
References......Page 327
Index......Page 340