Table 4.

Secondary outcome regression results using both patient-reported and administrative claims data at 12 months after myocardial infarction. Values are odds ratio (95% confidence interval)

Source, outcome, and intervention*	Effect estimate† (95% CI)	P value
Patient-reported‡ (n=2502)
Cardiac rehabilitation enrolment (yes/no):
Mail-outs§ (n=829)	1.27 (1.01 to 1.61)	0.05
Mail-outs plus phone calls§ (n=844)	1.55 (1.23 to 1.95)	<.001
Cardiac rehabilitation attendance (%, continuous):
Mail-outs§ (n=829)	4.33 (−0.60 to 9.27)	0.09
Mail-outs plus phone calls§ (n=844)	10.9 (5.10 to 16.7)	<.001
Adherence to medication in past 30 days (number of drug classes with <6 days missed; 0-4):
Mail-outs§ (n=829)	0.97 (0.81 to 1.17)	0.78
Mail-outs plus phone calls§ (n=844)	0.91 (0.75 to 1.12)	0.38
Adherence to medication to in past 7 days (yes/no):
Mail-outs§ (n=829)	0.97 (0.76 to 1.23)	0.79
Mail-outs plus phone calls§ (n=844)	0.86 (0.68 to 1.10)	0.24
Statin adherence in past 7 days (yes/no):
Mail-outs§ (n=829)	1.02 (0.78 to 1.32)	0.91
Mail-outs plus phone calls§ (n=844)	0.95 (0.73 to 1.24)	0.73
Antiplatelet adherence in past 7 days (yes/no):
Mail-outs§ (n=829)	0.85 (0.58 to 1.26)	0.41
Mail-outs plus phone calls§ (n=844)	0.82 (0.58 to 1.16)	0.25
β blocker adherence in past 7 days (yes/no):
Mail-outs§ (n=829)	1.02 (0.80 to 1.30)	0.88
Mail-outs plus phone calls§ (n=844)	1.00 (0.79 to 1.27)	0.99
Angiotensin system inhibitor adherence in past 7 days (yes/no):
Mail-outs§ (n=829)	1.08 (0.85 to 1.38)	0.51
Mail-outs plus phone calls§ (n=844)	1.08 (0.84 to 1.40)	0.53
Persistence with medication in all four classes (yes/no):
Mail-outs§ (n=829)	1.00 (0.79 to 1.25)	0.97
Mail-outs plus phone calls§ (n=844)	0.96 (0.76 to 1.21)	0.71
Persistence with statins (yes/no):
Mail-outs§ (n=829)	1.00 (0.72 to 1.40)	>.999
Mail-outs plus phone calls§ (n=844)	1.00 (0.75 to 1.32)	0.99
Persistence with antiplatelets (yes/no):
Mail-outs§ (n=829)	0.77 (0.50 to 1.18)	0.23
Mail-outs plus phone calls§ (n=844)	0.86 (0.57 to 1.31)	0.49
Persistence with β blocker (yes/no):
Mail-outs§ (n=829)	1.01 (0.77 to 1.31)	0.96
Mail-outs plus phone calls§ (n=844)	1.09 (0.82 to 1.46)	0.54
Persistence with angiotensin system inhibitor (yes/no):
Mail-outs§ (n=829)	1.04 (0.83 to 1.31)	0.74
Mail-outs plus phone calls§ (n=844)	1.08 (0.84 to 1.40)	0.53
Adherence to dual platelets in past 7 days (yes/no):
Mail-outs§ (n=829)	0.97 (0.78 to 1.20)	0.76
Mail-outs plus phone calls§ (n=844)	1.00 (0.80 to 1.26)	0.97
Quality of life (continuous):
Mail-outs§ (n=829)	0.09 (−1.38 to 1.54)	0.91
Mail-outs plus phone calls§ (n=844)	−1.00 (−2.73 to 0.72)	0.25
Smoking status (yes/no):
Mail-outs§ (n=829)	1.02 (0.73 to 1.42)	0.91
Mail-outs plus phone calls§ (n=844)	1.03 (0.75 to 1.40)	0.87
Administrative claims¶ (n=1406)
Adherence to medication†† – PDC ≥80% (yes/no):
Mail-outs§ (n=482)	1.14 (0.87 to 1.49)	0.35
Mail-outs plus phone calls§ (n=450)	0.99 (0.75 to 1.30)	0.95
Adherence to statins – PDC ≥80% (yes/no):
Mail-outs§ (n=482)	0.89 (0.69 to 1.16)	0.39
Mail-outs plus phone calls§ (n=450)	1.04 (0.80 to 1.36)	0.78
Adherence to antiplatelets†† – PDC ≥80% (yes/no):
Mail-outs§ (n=482)	1.09 (0.85 to 1.41)	0.51
Mail-outs plus phone calls§ (n=450)	1.10 (0.85 to 1.43)	0.48
Adherence to β blocker – PDC ≥80% (yes/no):
Mail-outs§ (n=482)	0.91 (0.70 to 1.17)	0.48
Mail-outs plus phone calls§ (n=450)	0.99 (0.76 to 1.29)	0.95
Adherence to angiotensin system inhibitor – PDC ≥80% (yes/no):
Mail-outs§ (n=482)	1.02 (0.79 to 1.31)	0.89
Mail-outs plus phone calls§ (n=450)	0.95 (0.73 to 1.23)	0.71
Administrative claims** (n=2624)
No of outpatient visits:
Mail-outs§ (n=874)	1.03 (0.97 to 1.10)	0.36
Mail-outs plus phone calls§ (n=875)	1.03 (0.97 to 1.09)	0.33
No of emergency department visits without admission:
Mail-outs§ (n=874)	1.27 (1.08 to 1.48)	0.003
Mail-outs plus phone calls§ (n=875)	1.08 (0.92 to 1.26)	0.34
No of hospital admissions:
Mail-outs§ (n=874)	1.05 (0.90 to 1.23)	0.55
Mail-outs plus phone calls§ (n=875)	1.02 (0.87 to 1.19)	0.82
Repeat myocardial infarction, stroke, or coronary revascularisation (yes/no):
Mail-outs§ (n=874)	0.96 (0.73 to 1.27)	0.79
Mail-outs plus phone calls§ (n=875)	0.89 (0.68 to 1.18)	0.41
All cause mortality‡‡:
Mail-outs§ (n=874)	0.98 (0.67 to 1.45)	0.93
Mail-outs plus phone calls§ (n=875)	0.67 (0.43 to 1.03)	0.07

PDC=proportion of days covered. All effect estimates are adjusted for fixed effect of centre (stratifying factor in randomisation). Fully conditional specification was used to create 20 imputed datasets for each outcome (that is, multiple imputation). These multiple datasets were then analysed independently using regression analysis. The effect estimates and 95% confidence intervals presented were obtained by pooling regression results across the imputed datasets using Rubin’s rules (see statistical appendix for more details).

^*Reference group=usual care.

^†Effect estimate is an absolute mean difference for continuous outcomes (β=MD), odds ratio for categorical outcomes (exp(β)=OR), rate ratio for count-based outcomes (exp(β)=RR), and hazard ratio (exp(β)=HR) for all cause mortality (which was modelled as a time-to-event outcome).

^‡Excludes individuals who died between 28 and 365 days after randomisation.

^§Assumes proportionality of effect across all four logits (4 v <4, ≥3 v <3, ≥2 v <2, ≥1 v 0).

^¶Limited to 1412 older adults (65 and older) in initial sample. Six participants not used in sample owing to inability to link their patient reported data with administrative claims data.

^**Smaller number than the patient reported sample for co-primary outcomes (n=2632) due to inability to link eight individuals to administrative claims data.

^††Excluding aspirin from antiplatelet drug classes owing to inadequate capture in administrative claims data.

^‡‡Occurrence of death and death date obtained from administrative (that is, Ontario Registered Persons Database) claims.