Electronic Supplementary Material 1: Strong ties versus weak ties between GPs and SPs: sensitivity analysis forpatient characteristics

Patient characteristics / Strong ties (uppermost decile) / Weak ties (lowest decile) / p-value / Strong ties (uppermost tertile) / Weak ties (lowest tertile) / p-value
Gender (%)
male / 53.24 / 53.60 / 0.875 / 53.03 / 51.86 / 0.321
female / 46.76 / 46.40 / 46.97 / 48.14
Average age / 66.08 / 64.81 / 0.000 / 66.03 / 64.40 / 0.000
Diabetes (% based on ICD-10 codes)5
with complications / 57.37 / 61.26 / 0.083 / 56.99 / 58.25 / 0.277
without complications / 42.63 / 38.74 / 43.01 / 41.75
Treatment (%based on third-level ATC codes)
insulin based / 68.61 / 72.76 / 0.048 / 68.96 / 72.90 / 0.000
non-insulin based / 31.39 / 27.24 / 31.04 / 27.10
Prescriptions per patient
for all agents / 131.90 / 142.50 / 0.017 / 132.90 / 125.53 / 0.001
for antidiabetic agents / 27.14 / 28.15 / 0.229 / 27.42 / 25.58 / 0.000
Consultations per patient
with the GP for all prescriptions / 22.71 / 21.99 / 0.255 / 22.72 / 21.00 / 0.000
with the SP for therapy-initiating prescriptions / 2.32 / 2.92 / 0.000 / 2.34 / 2.46 / 0.029

Electronic Supplementary Material 2A: Prevalence of comorbid conditions included in the Charlson Comorbidity Index forvarious diabetes cohorts (%)

Charlson Comorbidity Index / Lubloy et al. / Meyers et al. (2014) / Lix et al. (2013)
Study cohort, n=31,070 / Upper cost group
(top 20%), n=344,019 / Lower cost group
(bottom 80%), n=1,376,022 / Cohort1,
n=29,058 / Cohort2, n=41,925
AIDS/HIV / 0.0 / 0.5 / 0.1 / <0.1 / <0.1
Any malignancy, including lymphoma, and leukemia, except malignant neoplasm of skin / 2.4 / 15.7 / 5.9 / 5.5 / 5.8
Cerebrovascular disease / 23.2 / 15.6 / 6.3 / 5.5 / 4.7
Chronic pulmonary disease / 13.3 / 26.1 / 12.6 / 14.0 / 13.7
Congestive heart failure / 15.5 / 16.2 / 4.7 / 8.0 / 8.1
Dementia / 0.9 / 1.4 / 0.7 / 0.8 / 0.9
Hemiplegia or paraplegia / 0.2 / 1.5 / 0.2 / 0.6 / 0.4
Metastatic solid tumor / 0.3 / 3.6 / 0.4 / 1.3 / 1.8
Mild liver disease / 2.0 / 1.6 / 0.3 / 0.6 / 0.6
Moderate or severe liver disease / 0.3 / 19.2 / 6.3 / 0.1 / 0.2
Myocardial infarction / 4.2 / 8.4 / 1.7 / 2.5 / 3.0
Peptic ulcer disease / 4.3 / 2.5 / 0.7 / 2.0 / 1.7
Peripheral vascular disease / 18.5 / 10.0 / 4.0 / 1.4 / 1.6
Renal disease / 1.5 / 7.6 / 1.5 / 1.9 / 3.2
Rheumatic disease / 1.5 / 4.4 / 1.8 / 0.8 / 0.8

Electronic Supplementary Material 2B: Prevalence of comorbid conditions included in the Elixhauser measure forvarious diabetes cohorts (%)

Elixhauser Comorbidity Index / Lubloy et al. / Lix et al. (2013) / Quail et al. (2011)
Study cohort, n=31,070 / Cohort1,n=29,058 / Cohort2, n=41,925 / Full cohort, n=662,423 / 65+ years, n=41,925
AIDS/HIV / 0.0 / <0.1 / <0.1 / <0.1 / 0.0
Alcohol abuse / 0.5 / 0.4 / 0.7 / 0.7 / 0.4
Blood loss anemia / 1.1 / 0.0 / <0.1 / 0.1 / 0.1
Cardiac arrhythmias / 9.3 / 2.3 / 2.8 / 2.8 / 4.7
Chronic pulmonary disease / 13.2 / 13.9 / 13.5 / 13.5 / 14.3
Coagulopathy / 0.4 / 1.0 / 2.5 / 2.5 / 3.9
Congestive heart failure / 15.5 / 8.0 / 8.0 / 8.0 / 13.4
Deficiency anemia / 4.4 / 0.5 / 0.8 / 0.8 / 1.3
Depression / 20.1 / 5.8 / 6.0 / 6.0 / 4.8
Drug abuse / 0.0 / 0.3 / 0.4 / 0.4 / 0.1
Fluid and electrolyte disorders / 4.4 / 1.9 / 2.7 / 2.7 / 4.0
Hypertension, complicated / 2.7 / 0.7 / 1.0 / 1.0 / 1.5
Hypertension, uncomplicated / 89.4 / 33.3 / 42.7 / 42.7 / 51.1
Hypothyroidism / 2.9 / 0.4 / 1.1 / 4.7 / 5.6
Liver disease / 2.2 / 0.6 / 0.7 / 0.7 / 0.6
Lymphoma / 0.2 / 0.2 / 0.2 / 0.2 / 0.3
Metastatic cancer / 0.3 / 1.3 / 1.7 / 1.7 / 2.8
Obesity / 0.4 / 0.6 / 1.0 / 1.0 / 0.9
Other neurological disorders / 3.5 / 1.1 / 1.2 / 1.2 / 1.3
Paralysis / 0.2 / 0.6 / 0.4 / 0.4 / 0.6
Peptic ulcer disease excluding bleeding / 3.1 / 0.3 / 0.4 / 0.4 / 0.5
Peripheral vascular disorders / 18.5 / 1.4 / 1.5 / 1.5 / 2.2
Psychoses / 0.4 / 2.4 / 2.7 / 2.7 / 4.4
Pulmonary circulation disorders / 0.1 / 0.7 / 0.8 / 0.8 / 1.2
Renal failure / 1.5 / 1.8 / 3.1 / 3.1 / 4.2
Rheumatoid arthritis/ collagen vascular diseases / 1.6 / 1.9 / 1.9 / 1.9 / 2.2
Solid tumor without metastasis / 2.1 / 5.0 / 5.4 / 5.4 / 8.8
Valvular disease / 0.7 / 1.5 / 2.2 / 2.2 / 3.1
Weight loss / 0.4 / 0.1 / 0.2 / 0.2 / 0.2

Electronic Supplementary Material 3: Strong ties versus weak ties between GPs and SPs: sensitivity analysis foroutcome measures

Outcome measures / Strong ties (uppermost decile) / Weak ties (lowest decile) / p-value / Strong ties (upper-most tertile) / Weak ties (lowest tertile) / p-value
Patient health status (excluding diabetes)
Charlson comorbidity index / 0.93 / 0.97 / 0.522 / 0.94 / 0.93 / 0.888
Quan-modified Charlson comorbidity index / 0.60 / 0.63 / 0.573 / 0.61 / 0.62 / 0.683
Elixhauser measure (based on ICD-10 codes) / 1.98 / 1.95 / 0.716 / 1.99 / 1.94 / 0.133
ATC-based comorbidity count (based on third-level ATC codes) / 7.99 / 8.30 / 0.164 / 8.03 / 7.79 / 0.033
Pharmacy costs (thousand HUF) / 584.65 / 771.55 / 0.000 / 616.60 / 661.16 / 0.001

Electronic Supplementary Material 4: Determinants of pharmacy costs: multivariate regression analysis on tie strength and patient characteristics*

Independent variables / Standardized Coefficients / Sig.
Beta
Strong ties versus weak ties between GPs and SPs (0-strong, 1-weak) / 0.011 / 0.030
Quan-modified Charlson comorbidity index / -0.009 / 0.113
Patients' age / 0.174 / 0.001
Patients' agesquared / -0.272 / 0.000
Patients’ gender (0-female, 1-male) / -0.024 / 0.000
Diabetes without complications/ with complications (0/1) / 0.014 / 0.007
Treatment method (0 - non-insulin based, 1-insulin based) / 0.202 / 0.000
Prescriptions per patient for all agents / 0.570 / 0.000
Consultations per patient (with the SP for therapy-initiating prescriptions) / 0.062 / 0.000
R2=.396

* Variables excluded from the multivariate regression analysis due to high correlation (<0.6): Charlson comorbidity index, Elixhauser measure, ATC-based comorbidity count, Prescriptions per patient for antidiabetic agents, Consultations per patient with the GP.

------

Note: This is supplementary material for the manuscript entitled ‘Formal professional relationships between general practitioners and specialists in shared care: possible associations with patient health and pharmacy costs’ and is for online publication only.

Authors: Ágnes Lublóy, Judit Lilla Keresztúri, and Gábor Benedek

1