ONLINE SUPPLEMENTARY MATERIAL
Quality of life after TIA & stroke: 10-year results of the Oxford Vascular Study
Ramon Luengo-Fernandez DPhil, Alastair M. Gray PhD, Linda Bull RGN, Sarah Welch RGN, Fiona Cuthbertson MCSP, Peter M. Rothwell FMedSci forthe Oxford Vascular Study
Supplemental Methods
Multiple imputation
Multiple imputation is a statistical technique for handling missing data. This method uses the distribution of theobserved data to estimate a set of plausible values for the missing data (Sterne et al. 2009), withrandom components being incorporated into theseestimated values to reflect their uncertainty (i.e. we can never know the true values of the missing data) (White et al. 2010). The multiple imputation analysis for this analysis was undertaken using the “mi impute” command in STATA v. 12.
Given that we had missing information in a number of variables (Table e-1) our multiple imputation analysis was undertaken using chained equations. In this analysis all missing values were initially filled in by simple random sampling with replacement from the observed values. The first variable with missing values, e.g. NIHSS score, was regressed on all other explanatory variables, restrictedto individuals with observed NIHSS scores. Missing values in NIHSS scores were then replaced by simulated draws from the results of the regression model for NIHSS score. Then, the next variable with missing values, e.g. utility at 1 month, was regressed on all othervariables, restricted to individuals with observed utility values at 1 month, and using the imputed values of NIHSS score. Again, missingvalues in utility at 1 month were replaced by draws from the results of the regression model for utility. Details of the regression equations used in the multiple imputation analysis are reported in Table e-2. The process was repeated for all othervariables with missing values in turn (i.e.one cycle). This procedure was then repeated for 10 cycles to produce a single imputed data set, with a total of 10 imputed data sets being generated.
References:
Sterne AC, White IR, Carlin JB, Spratt M, Royston P, Kenward MG, Wood AM, Carpenter JR. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ 2009;338:b2393.
White IR, Royston P, Wood AM. Multiple imputation using chained equations: Issues and guidance for practice. Statist Med 2011;30:377-399.
Table e-1. Number of complete cases and cases with missing data
Variable / Complete / Incomplete & to impute(% missing) / Total cases
Age at time of index event / 1188 / 0 / 1188
Gender / 1188 / 0 / 1188
Pre-morbid mRS score / 1188 / 0 / 1188
Previous history of:
Hypertension / 1183 / 5 (<1) / 1188
Stroke / 1183 / 5 (<1) / 1188
Atrial fibrillation / 1183 / 5 (<1) / 1188
Angina / 1183 / 5 (<1) / 1188
Diabetes / 1183 / 5 (<1) / 1188
Peripheral Vascular Disease / 1183 / 5 (<1) / 1188
NIHSS score / 1176 / 12 (1) / 1188
Event type (Stroke or TIA) / 1188 / 0 / 1188
Recurrent events:
TIA / 1188 / 0 / 1188
Stroke / 1188 / 0 / 1188
Coronary / 1188 / 0 / 1188
Lived alone / 1091 / 97 (8) / 1188
Socio-economic status / 1060 / 101 (8) / 1188
Marital status / 1083 / 76 (6) / 1188
Age left education / 1065 / 97 (8) / 1188
Deprivation index / 1188 / 0 / 1188
mRS scores* at:
1 month / 1175 / 13 (1) / 1188
6 months / 1156 / 32 (3) / 1188
12 months / 1143 / 45 (4) / 1188
24 months / 813 / 375 (32) / 1188
60 months / 1065 / 123 (10) / 1188
EQ-5D utility† at:
1 month / 870 / 318 (27) / 1188
6 months / 761 / 427 (36) / 1188
12 months / 945 / 243 (20) / 1188
24 months / 716 / 472 (40) / 1188
60 months / 949 / 239 (20) / 1188
* mRS scores ranged from 0 (no symptoms) to 6 (death), with 5 representing severe disability.
†Given that in the EQ-5D tariff a utility value 0 represents death, cases in which the patient was dead a utility value of 0 was applied.
Table e-2. Regression equations used in multiple imputation of missing data
Variable(type, no. categories) / Regression Type / Explanatory variables
Hypertension history
(categorical, 2) / Logistic / Age; Gender; Pre-morbid mRS; History of each of - Stroke, Atrial fibrillation, Myocardial infarction, Angina, Diabetes, PVD; NIHSS score; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Lived alone; Socio-economic status; Marital status; Age left education; Deprivation index; Utility at 1, 6, 12, 24, 60 months; mRS at 1, 6, 12, 24, 60 months.
Stroke history
(categorical, 2) / Logistic / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Atrial fibrillation, Myocardial infarction, Angina, Diabetes, PVD; NIHSS score; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Lived alone; Socio-economic status; Marital status; Age left education; Deprivation index; Utility at 1, 6, 12, 24, 60 months; mRS at 1, 6, 12, 24, 60 months.
Atrial fibrillation history
(categorical, 2) / Logistic / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Myocardial infarction, Angina, Diabetes, PVD; NIHSS score; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Lived alone; Socio-economic status; Marital status; Age left education; Deprivation index; Utility at 1, 6, 12, 24, 60 months; mRS at 1, 6, 12, 24, 60 months.
Myocardial infarction history
(categorical, 2) / Logistic / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Atrial fibrillation, Angina, Diabetes, PVD; NIHSS score; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Lived alone; Socio-economic status; Marital status; Age left education; Deprivation index; Utility at 1, 6, 12, 24, 60 months; mRS at 1, 6, 12, 24, 60 months.
Angina history
(categorical, 2) / Logistic / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Atrial fibrillation, Myocardial infarction, Diabetes, PVD; NIHSS score; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Lived alone; Socio-economic status; Marital status; Age left education; Deprivation index; Utility at 1, 6, 12, 24, 60 months; mRS at 1, 6, 12, 24, 60 months.
Diabetes history
(categorical, 2) / Logistic / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Atrial fibrillation, Myocardial infarction, angina, PVD; NIHSS score; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Lived alone; Socio-economic status; Marital status; Age left education; Deprivation index; Utility at 1, 6, 12, 24, 60 months; mRS at 1, 6, 12, 24, 60 months.
PVD history
(categorical, 2) / Logistic / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Atrial fibrillation, Myocardial infarction, Angina, Diabetes; NIHSS score; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Lived alone; Socio-economic status; Marital status; Age left education; Deprivation index; Utility at 1, 6, 12, 24, 60 months; mRS at 1, 6, 12, 24, 60 months.
Event severity, NIHSS score
(continuous) / Predictive mean matching / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Atrial fibrillation, Myocardial infarction, angina, Diabetes, PVD; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Lived alone; Socio-economic status; Marital status; Age left education; Deprivation index; Utility at 1, 6, 12, 24, 60 months; mRS at 1, 6, 12, 24, 60 months.
Lived alone before event
(categorical, 2) / Logistic / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Atrial fibrillation, Myocardial infarction, Angina, Diabetes, PVD; NIHSS score; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Socio-economic status; Marital status; Age left education; Deprivation index; Utility at 1, 6, 12, 24, 60 months; mRS at 1, 6, 12, 24, 60 months.
Socio-economic status
(categorical, 6) / Multinomial logistic / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Atrial fibrillation, Myocardial infarction, Angina, Diabetes, PVD; Lived alone; Marital status; Age left education; Deprivation index.
Marital status
(categorical, 5) / Multinomial logistic / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Atrial fibrillation, Myocardial infarction, Angina, Diabetes, PVD; Lived alone; Socio-economic status; Age left education; Deprivation index.
Age left education
(continuous) / Predictive mean matching / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Atrial fibrillation, Myocardial infarction, Angina, Diabetes, PVD; NIHSS score; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Lived alone; Socio-economic status; Marital status; Deprivation index; Utility at 1, 6, 12, 24, 60 months; mRS at 1, 6, 12, 24, 60 months.
EQ-5D utility at 1 month
(continuous) / Predictive mean matching / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Atrial fibrillation, Myocardial infarction, Angina, Diabetes, PVD; NIHSS score; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Lived alone; Socio-economic status; Marital status; Age left education; Deprivation index; Utility at 6, 12, 24, 60 months; mRS at 1, 6, 12, 24, 60 months.
EQ-5D utility at 6 month
(continuous) / Predictive mean matching / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Atrial fibrillation, Myocardial infarction, Angina, Diabetes, PVD; NIHSS score; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Lived alone; Socio-economic status; Marital status; Age left education; Deprivation index; Utility at 1, 12, 24, 60 months; mRS at 1, 6, 12, 24, 60 months.
EQ-5D utility at 1 year
(continuous) / Predictive mean matching / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Atrial fibrillation, Myocardial infarction, Angina, Diabetes, PVD; NIHSS score; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Lived alone; Socio-economic status; Marital status; Age left education; Deprivation index; Utility at 1, 6, 24, 60 months; mRS at 1, 6, 12, 24, 60 months.
EQ-5D utility at 2 years
(continuous) / Predictive mean matching / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Atrial fibrillation, Myocardial infarction, Angina, Diabetes, PVD; NIHSS score; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Lived alone; Socio-economic status; Marital status; Age left education; Deprivation index; Utility at 1, 6, 12, 60 months; mRS at 1, 6, 12, 24, 60 months.
EQ-5D utility at 5 years
(continuous) / Predictive mean matching / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Atrial fibrillation, Myocardial infarction, Angina, Diabetes, PVD; NIHSS score; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Lived alone; Socio-economic status; Marital status; Age left education; Deprivation index; Utility at 1, 6, 12, 24 months; mRS at 1, 6, 12, 24, 60 months.
mRS at 1 month
(categorical, 7) / Ordinal logistic / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Atrial fibrillation, Myocardial infarction, Angina, Diabetes, PVD; NIHSS score; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Lived alone; Socio-economic status; Marital status; Age left education; Deprivation index; Utility at 1, 6, 12, 24, 60 months; mRS at 6, 12, 24, 60 months.
mRS at 6 months
(categorical, 7) / Ordinal logistic / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Atrial fibrillation, Myocardial infarction, Angina, Diabetes, PVD; NIHSS score; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Lived alone; Socio-economic status; Marital status; Age left education; Deprivation index; Utility at 1, 6, 12, 24, 60 months; mRS at 1, 12, 24, 60 months.
mRS at 1 year
(categorical, 7) / Ordinal logistic / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Atrial fibrillation, Myocardial infarction, Angina, Diabetes, PVD; NIHSS score; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Lived alone; Socio-economic status; Marital status; Age left education; Deprivation index; Utility at 1, 6, 12, 24, 60 months; mRS at 1, 6, 24, 60 months.
mRS at 2 years
(categorical, 7) / Ordinal logistic / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Atrial fibrillation, Myocardial infarction, Angina, Diabetes, PVD; NIHSS score; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Lived alone; Socio-economic status; Marital status; Age left education; Deprivation index; Utility at 1, 6, 12, 24, 60 months; mRS at 1, 6, 12, 60 months.
mRS at 5 years
(categorical, 7) / Ordinal logistic / Age; Gender; Pre-morbid mRS; History of each of - Hypertension, Stroke, Atrial fibrillation, Myocardial infarction, Angina, Diabetes, PVD; NIHSS score; Event type; Recurrent events during follow-up – TIA, stroke, coronary; Lived alone; Socio-economic status; Marital status; Age left education; Deprivation index; Utility at 1, 6, 12, 24, 60 months; mRS at 1, 6, 12, 24 months.
Supplemental Results
Tablee-3. Reasons for missing EQ-5D information
1 month / 6 months / 1 year / 2 years / 5 yearsAs proportion (%) of total number of patients
Dead / 111 (9) / 178 (15) / 222 (19) / 280 (24) / 470 (40)
Alive / 1, 077 (91) / 1, 010 (85) / 966 (81) / 908 (76) / 718 (60)
As proportion (%) of total number of survivors
Cases with EQ-5D data / 759 (70) / 583 (58) / 723 (75) / 436 (48) / 479 (67)
Cases without EQ-5D data / 318 (30) / 427 (42) / 243 (25) / 472 (52) / 239 (33)
As proportion (%) of total number of cases with missing EQ-5D data
Reasons for no EQ-5D information
EQ-5D not provided at 6m / N/A / 251 (59) / N/A / N/A / N/A
2y follow-up discontinued / N/A / N/A / N/A / 338 (72) / N/A
Late referral/ascertainment / 58 (18) / 10 (2) / 2 (1) / 0 / 0
Too unwell / 37 (11) / 13 (3) / 19 (8) / 11 (2) / 16 (6)
Dysphasia / 20 (6) / 3 (1) / 8 (3) / 1 (<1) / 5 (2)
Dementia / 10 (3) / 9 (2) / 12 (5) / 10 (2) / 15 (6)
Other health problem / 11 (3) / 7 (2) / 9 (4) / 7 (1) / 17 (7)
No English/Illiterate / 6 (2) / 1 (<1) / 3 (1) / 3 (1) / 4 (2)
Telephone follow-up* / 46 (14) / 30 (7) / 55 (23) / 50 (11) / 60 (25)
Refused follow-up / 26 (8) / 33 (8) / 46 (19) / 25 (5) / 41 (17)
Refused to/did not complete EQ-5D / 41 (13) / 6 (1) / 7 (3) / 5 (1) / 18 (7)
Lost to follow-up/DNA† / 63 (20) / 64 (15) / 82 (34) / 22 (5) / 63 (26)
* Telephone follow-ups were undertaken for patients who either moved out of area, or for whom a hospital/home visit was not possible to arrange. Telephone follow-ups did not include the EQ-5D.
†DNA-Did not attend scheduled follow-up
Tablee-4. EQ-5D responses after index TIA or stroke
TIA / Stroke
Mobility / Self-care / Activities / Pain / Anxiety / Mobility / Self-care / Activities / Pain / Anxiety
1 month
No / 186 (59) / 269 (85) / 232 (73) / 198 (63) / 223 (71) / 191 (42) / 290 (64) / 170 (38) / 272 (61) / 279 (62)
Some / 127 (40) / 43 (14) / 74 (23) / 105 (33) / 83 (26) / 223 (49) / 129 (29) / 183 (41) / 155 (35) / 146 (33)
Extreme / 3 (1) / 3 (1) / 10 (3) / 12 (4) / 9 (3) / 37 (8) / 32 (7) / 98 (22) / 20 (4) / 22 (5)
No. cases / 316 / 315 / 316 / 315 / 315 / 451 / 451 / 451 / 447 / 447
6 months
No / 139 (57) / 209 (86) / 171 (70) / 146 (60) / 169 (69) / 168 (49) / 239 (70) / 170 (50) / 204 (60) / 239 (70)
Some / 103 (42) / 32 (13) / 61 (25) / 84 (34) / 66 (27) / 161 (47) / 84 (25) / 132 (39) / 120 (35) / 91 (27)
Extreme / 2 (1) / 3 (1) / 12 (5) / 14 (6) / 9 (4) / 12 (4) / 17 (5) / 37 (11) / 16 (5) / 10 (3)
No. cases / 242 / 244 / 244 / 244 / 244 / 341 / 340 / 339 / 340 / 340
1 year
No / 178 (58) / 254 (83) / 204 (67) / 195 (64) / 232 (76) / 171 (41) / 283 (68) / 198 (47) / 236 (56) / 287 (69)
Some / 127 (42) / 48 (16) / 85 (28) / 99 (32) / 66 (22) / 243 (58) / 123 (29) / 178 (42) / 164 (39) / 119 (28)
Extreme / 1 (<1) / 4 (1) / 17 (6) / 12 (4) / 7 (2) / 5 (1) / 13 (3) / 43 (10) / 18 (4) / 12 (3)
No. cases / 306 / 306 / 306 / 306 / 305 / 419 / 419 / 419 / 418 / 418
2 years
No / 92 (53) / 142 (82) / 110 (63) / 105 (60) / 124 (72) / 106 (40) / 175 (67) / 120 (46) / 142 (54) / 175 (67)
Some / 81 (47) / 28 (16) / 53 (30) / 62 (36) / 44 (25) / 145 (55) / 72 (27) / 111 (42) / 110 (42) / 81 (31)
Extreme / 1 (1) / 4 (2) / 11 (6) / 7 (4) / 5 (3) / 12 (5) / 16 (6) / 32 (12) / 11 (4) / 7 (3)
No. cases / 174 / 174 / 174 / 174 / 173 / 263 / 263 / 263 / 263 / 263
5 years
No / 124 (59) / 179 (85) / 153 (73) / 118 (56) / 162 (77) / 125 (46) / 192 (71) / 162 (60) / 128 (48) / 196 (73)
Some / 85 (40) / 29 (14) / 52 (25) / 89 (42) / 46 (22) / 136 (51) / 68 (25) / 83 (31) / 122 (45) / 59 (22)
Extreme / 1 (<1) / 3 (1) / 5 (2) / 4 (2) / 2 (1) / 8 (3) / 9 (3) / 24 (9) / 19 (7) / 14 (5)
No. cases / 210 / 211 / 210 / 211 / 210 / 269 / 269 / 269 / 269 / 269
Table e-5.Discounted 5-year life expectancy and quality-adjusted life expectancy after index stroke or TIA
5-year life expectancyMean life years (95% CI) / 5-year quality-adjusted life expectancy
Mean QALYs (95% CI)
3.0% / 3.5% / 5.0% / 3.0% / 3.5% / 5.0%
TIA / 4.02 (3.93, 4.13) / 3.97 (3.89, 4.08) / 3.85 (3.77, 3.95) / 2.88 (2.80, 3.02) / 2.82 (2.74, 2.95) / 2.63 (2.56, 2.75)
Stroke / 3.08 (3.01, 3.24) / 3.05 (2.98, 3.21) / 2.95 (2.89, 3.11) / 1.93 (1.88, 2.07) / 1.88 (1.84, 2.02) / 1.76 (1.72, 1.89)
Stroke severity
Minor / 3.79 (3.67, 3.92) / 3.75 (3.63, 3.88) / 3.63 (3.52, 3.75) / 2.56 (1.45, 2.67) / 2.51 (2.39, 2.61) / 2.34 (2.24, 2.44)
Moderate / 2.59 (2.51, 3.03) / 2.56 (2.48, 3.00) / 2.49 (2.41, 2.91) / 1.44 (1.35, 1.71) / 1.41 (1.32, 1.68) / 1.32 (1.24, 1.57)
Severe / 1.31 (1.14, 1.69) / 1.30 (1.13, 1.67) / 1.27 (1.10, 1.63) / 0.61 (0.36, 0.68) / 0.60 (0.35, 0.67) / 0.56 (0.33, 0.62)
Stroke type
Ischaemic / 3.38 (3.31, 3.42) / 3.34 (3.28, 3.38) / 3.24 (3.18, 3.28) / 2.16 (2.04, 2.18) / 2.11 (2.00, 2.13) / 1.97 (1.87, 2.00)
Primary intracerebral haemorrhage / 2.20 (1.97, 2.43) / 2.18 (1.96, 2.41) / 2.12 (1.91, 2.35) / 1.52 (1.29, 1.73) / 1.49 (1.26, 1.69) / 1.38 (1.17, 1.58)
Subarachnoid haemorrhage / 2.52 (2.48, 2.53) / 2.51 (2.47, 2.52) / 2.49 (2.45, 2.50) / 1.86 (1.63, 1.93) / 1.81 (1.59, 1.89) / 1.69 (1.48, 1.76)
Unknown stroke / 1.44 (1.05, 1.51) / 1.43 (1.05, 1.50) / 1.40 (1.02, 1.47) / 0.68 (0.46, 0.80) / 0.66 (0.45, 0.77) / 0.63 (0.43, 0.73)
Stroke (TOAST classification)
Cardioembolism / 2.45 (2.36, 2.57) / 2.42 (2.34, 2.55) / 2.35 (2.27, 2.48) / 1.44 (1.35, 1.59) / 1.41 (1.32, 1.56) / 1.32 (1.24, 1.46)
Large artery artherosclerosis / 3.50 (3.41, 3.73) / 3.47 (3.38, 3.70) / 3.37 (3.28, 3.60) / 1.98 (1.96, 2.39) / 1.94 (1.92, 2.34) / 1.81 (1.79, 2.18)
Small vessel occlusion / 4.16 (3.49, 4.21) / 4.11 (3.45, 4.17) / 3.99 (3.33, 4.05) / 2.54 (2.39, 2.71) / 2.49 (2.33, 2.65) / 2.33 (2.19, 2.48)
Other determined/multiple aetiologies / 3.09 (2.79, 3.22) / 3.07 (2.73, 3.20) / 3.00 (2.67, 3.13) / 1.94 (1.47, 2.16) / 1.90 (1.44, 2.11) / 1.78 (1.35, 1.98)
Undertermined aetiology / 4.10 (3.78, 4.11) / 4.05 (3.74, 4.06) / 3.93 (3.62, 3.94) / 2.65 (2.53, 2.80) / 2.59 (2.47, 2.73) / 2.42 (2.31, 2.55)
Unknown aetiology / 1.95 (1.85, 2.17) / 1.94 (1.83, 2.15) / 1.89 (1.79, 2.10) / 0.97 (0.94, 1.21) / 0.95 (0.92, 1.18) / 0.89 (0.86, 1.11)
Country-specific economic evaluation guidelines in Finland, Germany, Italy, Spain, Sweden and the USA recommend discounting future health gains at 3.0%.
Those in England & Wales recommend discounting future health gains at 3.5%, and guidelines in Australia, Austria, Canada and Portugal recommend using a discount rate of 5.0%. French guidelines recommend discounting future health gains using a rate of 3% and 5%.
Reference: International Society for Pharmacoeconomics and Outcomes Research. Pharmacoeconomic Guidelines Around the World [online]. Accessed June 4 2013.
Table e-6. Regression equations to derive the 1-month and 5-year probability of reporting problems in the EQ-5D, and utility loss, conditional on reporting problems in the EQ-5D.
1 month / 5 yearsEquation to derive probability of reporting problems in EQ-5D* / Equation to derive utility loss, conditional on reporting problems in EQ-5D† / Equation to derive probability of reporting problems in EQ-5D** / Equation to derive utility loss, conditional on reporting problems in EQ-5D‡
Coeff. / p<|z| / Coeff. / p<|z| / Coeff. / p<|z| / Coeff. / p<|z|
Age / 0.009 / 0.289 / 0.041 / 0.041 / 0.041 / 0.001 / 0.004 / 0.188
Male / -0.870 / <0.001 / -0.525 / -0.525 / -0.525 / 0.051 / -0.223 / 0.002
Previous history of:
Disability / 1.592 / 0.001 / 1.107 / 1.107 / 1.107 / 0.172 / 0.289 / 0.057
Hypertension / 0.193 / 0.322 / 0.122 / 0.122 / 0.122 / 0.624 / 0.072 / 0.234
Stroke / 0.123 / 0.662 / 0.504 / 0.504 / 0.504 / 0.199 / 0.098 / 0.257
Atrial fibrillation / 0.341 / 0.203 / -0.072 / -0.072 / -0.072 / 0.856 / -0.072 / 0.335
Myocardial infarction / -0.402 / 0.235 / 1.041 / 1.041 / 1.041 / 0.053 / -0.196 / 0.048
Angina / -0.173 / 0.558 / -0.286 / -0.286 / -0.286 / 0.487 / 0.229 / 0.016
Diabetes / 0.598 / 0.070 / 0.651 / 0.651 / 0.651 / 0.137 / -0.027 / 0.769
Peripheral vascular disease / 1.019 / 0.024 / 0.709 / 0.709 / 0.709 / 0.237 / -0.006 / 0.955
NIHSS score / 0.161 / 0.002 / 0.243 / 0.243 / 0.243 / 0.006 / 0.043 / <0.001
Event type
TIA / Reference case / Reference case / Reference case / Reference case
Stroke / 0.537 / 0.014 / 0.315 / 0.315 / 0.315 / 0.241 / 0.108 / 0.141
Recurrent vascular events
TIA / 0.306 / 0.361 / 0.256 / 0.256 / 0.256 / 0.146 / 0.041 / 0.338
Stroke / 1.087 / 0.030 / 0.825 / 0.825 / 0.825 / 0.012 / 0.043 / 0.380
Coronary / -0.017 / 0.988 / 0.162 / 0.162 / 0.162 / 0.652 / 0.072 / 0.059
Lived alone before event / 0.203 / 0.513 / 0.081 / 0.081 / 0.081 / 0.862 / -0.262 / 0.010
Socio-economic status
Professional / 0.264 / 0.509 / 0.655 / 0.655 / 0.655 / 0.205 / -0.154 / 0.196
Managerial/technical / -0.003 / 0.992 / 0.397 / 0.397 / 0.397 / 0.263 / -0.084 / 0.429
Skilled non-manual / -0.229 / 0.433 / 0.457 / 0.457 / 0.457 / 0.207 / -0.182 / 0.052
Skilled manual / Reference case / Reference case / Reference case / Reference case
Partly skilled / 0.176 / 0.606 / 0.711 / 0.711 / 0.711 / 0.119 / -0.049 / 0.704
Unskilled / 0.100 / 0.791 / -0.272 / -0.272 / -0.272 / 0.532 / -0.112 / 0.347
Marital status
Married / Reference case / Reference case / Reference case / Reference case
Widowed / -0.149 / 0.652 / 0.113 / 0.113 / 0.113 / 0.818 / 0.325 / 0.002
Single / -0.010 / 0.982 / 0.413 / 0.413 / 0.413 / 0.534 / 0.426 / 0.013
Separated/divorced / -0.336 / 0.446 / 0.975 / 0.975 / 0.975 / 0.106 / 0.231 / 0.179
Living with partner / -0.255 / 0.674 / 0.185 / 0.185 / 0.185 / 0.797 / -0.022 / 0.851
Age left education / -0.059 / 0.041 / -0.068 / -0.068 / -0.068 / 0.075 / -0.019 / 0.023
IMD / 0.013 / 0.427 / -0.011 / -0.011 / -0.011 / 0.606 / 0.005 / 0.424
Constant / 0.269 / 0.771 / -2.117 / -2.117 / -2.117 / 0.083 / -1.267 / <0.001
IMD – Index of Multiple Deprivation
Model specifications were tested using Preigbon’s link test. For the generalised linear model, the distributional family and identity were assessed using the modified Park test and the Box-Cox test.
*n=750; †n=559
**n=469; ‡n=327
Table e-7.Predictors of EQ-5D utility at 1 month and 5 years months following index TIA or stroke – multiple imputation
Expected mean (95% CI) utility gain at 1 month / Expected mean (95% CI) utility gain at 5 yearsConstant* / 0.767 (0.730, 0.797) / 0.726 (0.668, 0.782)
Age (additional year of age) / -0.001 (-0.002, -0.001)‡ / -0.007 (-0.009, -0.005)†
Male (vs. female) / 0.051 (0.026, 0.076)† / 0.047 (0.009, 0.086)‡
Previous history of (vs. no history):
Disability / -0.214 (-0.262, -0.170)† / -0.246 (-0.334, -0.168)†
Hypertension / -0.033 (-0.059, -0.010)† / -0.025 (-0.064, 0.010)
Stroke / -0.005 (-0.039, 0.026) / -0.045 (-0.099, 0.010)
Atrial fibrillation / -0.050 (-0.085, -0.016)† / 0.002 (-0.051, 0.047)
Myocardial infarction / 0.035 (-0.012, 0.074) / -0.018 (-0.072, 0.040)
Angina / -0.004 (-0.040, 0.032) / -0.010 (-0.073, 0.042)
Diabetes / -0.027 (-0.065, 0.010) / -0.027 (-0.081, 0.024)
Peripheral vascular disease / -0.087 (-0.140, -0.038)† / -0.053 (-0.126, 0.010)
NIHSS score (unit increase in NIHSS score) / -0.023 (-0.027, -0.018)† / -0.026 (-0.035, -0.018)†
Stroke (vs. TIA) / -0.052 (-0.079, -0.024)† / -0.074 (-0.116, -0.034)†
Recurrent events (vs. no recurrent events)
TIA / 0.009 (-0.034, 0.048) / -0.024 (-0.057, -0.001)‡
Stroke / -0.148 (-0.201, -0.091)† / -0.089 (-0.126, -0.054)†
Coronary / -0.074 (-0.179, 0.077) / -0.055 (-0.105, -0.013)‡
Lived alone before event (vs. living with someone) / -0.027 (-0.059, 0.006) / 0.053 (0.013, 0.093)‡
Socio-economic status (vs. skilled manual)
Professional / -0.005 (-0.056, 0.047) / 0.016 (-0.050, 0.078)
Managerial/technical / 0.017 (-0.017, 0.048) / 0.023 (-0.029, 0.072)
Skilled non-manual / 0.003 (-0.031, 0.035) / -0.006 (-0.055, 0.041)
Partly skilled / -0.025 (-0.068, 0.017) / -0.016 (-0.088, 0.046)
Unskilled / 0.012 (-0.030, 0.051) / 0.037 (-0.021, 0.095)
Marital status (vs. married)
Widowed / 0.005 (-0.033, 0.040) / -0.055 (-0.117, 0.001)
Single / -0.006 (-0.061, 0.045) / -0.122 (-0.238, -0.011)‡
Separated/divorced / 0.061 (-0.003, 0.116) / -0.130 (-0.227, -0.044)†
Living with partner / -0.062 (-0.189, 0.061) / -0.006 (-0.129, 0.112)
Age left education (additional year of education) / 0.007 (0.003, 0.011)† / 0.010 (0.006, 0.017)†
Deprivation - IMD score (unit increase in IMD score) / -0.001 (-0.003, 0.001) / -0.001 (-0.004, 0.002)
*Constant set at: age of 74 years at time of index event, NIHSS score of 3.4, age left education of 16 years, age left education of 16 years, IMD deprivation score of 9.6, and all dummy variables set to their reference case (as defined in brackets in the table).
†Significant at p<0.01; ‡Significant at p<0.05.
Figure e-1. Relationship between EQ-5D utility and event severity at A) 1 month B) 5 years.
A
B