Non-random relations in drug use expressed as patterns comprising prescription and over-the-counter drugs in multimorbid elderly patients in primary care: Data of the exploratory analysis of the multicentre, observational cohort study MultiCare

Figures & data

Table 1. Distribution of 24,535 drugs according to anatomical main group (ATC 1st level) and the top 20 ATC 5th level drugs (sorted according to their ATC 1st level) of 3189 patients of the MultiCare cohort and their proportion within the 1891 female patients in Germany (multiple use possible) (2008–2009).

ATC 1st level	ATC 5th level drugs	Frequency	Proportion per total number of drugs [%]	Proportion per patient [%]	Frequency for female (Proportion per female [%])
Cardiovascular system		9257	37.7		5162 (55.8%)
	Simvastatin	1114		34.9	526 (47.2%)
	Hydrochlorothiazide	1106		34.7	654 (59.1%)
	Ramipril	695		21.8	347 (49.9%)
	Metoprolol	661		20.7	375 (56.7%)
	Bisoprolol	629		19.7	368 (58.5%)
	Amlodipine	466		14.6	258 (55.4%)
	Torasemide	376		11.8	203 (54.0%)
	Enalapril	330		10.3	188 (57.0%)
	Lisinopril	216		6.8	109 (50.5%)
Alimentary tract and metabolism		5006	20.4		3286 (65.6%)
	Magnesium	765		24.0	541 (70.7%)
	Calcium	548		17.2	454 (82.8%)
	Omeprazole	448		14.0	278 (62.1%)
	Metformin	436		13.7	220 (50.5%)
	Cholecalciferol	392		12.3	337 (86.0%)
Nervous system		2507	10.2		1766 (70.4%)
Blood and blood forming organs		1904	7.8		955 (50.2%)
	Acetylsalicylic acid	1134		35.6	573 (50.5%)
	Phenprocoumon	441		13.8	215 (48.8%)
Musculo-skeletal system		1819	7.4		1186 (65.2%)
	Allopurinol	402		12.6	166 (41.3%)
	Diclofenac	390		12.2	268 (68.7%)
	Ibuprofen	335		10.5	244 (72.8%)
Respiratory system		1361	5.5		770 (56.6%)
Systemic hormonal preparations		999	4.1		773 (77.4%)
	Levothyroxine	616		19.3	504 (81.8%)
Genito-urinary system and sex hormones		568	2.3		241 (42.4 %)
Sensory organs		488	2.0		309 (63.3%)
Dermatologicals		190	0.8		123 (64.7%)
Antineoplastic and immunomodulating agents		154	0.6		122 (72.7%)
Various		106	0.4		67 (63.2%)
Antiinfectives		99	0.4		76 (76.8%)
Herbal and homeopathic agents		72	0.3		54 (75.0%)
Antiparasitic products		5	0.02		5 (100.0 %)

Table 2. Loading of factors with eigenvalue ≥ 1 and cumulative percent for ATC 3rd level substances of 1298 male patients in Germany (2008–2009).

Male
		D-OPD^a	D-CDH^b	D-Osteo^c	D-HF^d	D-Pain^e
	Eigenvalue	3.60	3.32	2.59	2.52	1.92
	Cumulative percent [%]	7.61	14.99	22.17	29.22	35.36
ATC 3^rd level		Factorloading
A02B	Drugs for peptic ulcer and gastro-oesophageal reflux diseases					0.60
A10A	Insulins and analogues	−0.32		0.31
A11C	Vitamin A und D, including combinations of the two			0.83
A12A	Calcium			0.88
A12C	Other mineral supplements			0.51		−0.34
B01A	Antithrombotic agents		0.85
C01A	Cardiac glycosides				0.89
C01D	Vasodilators used in cardiac diseases		0.46
C03C	High-ceiling diuretics		0.29	0.32	0.47
C03D	Potassium-sparing agents				0.34
C07A	Beta blocking agents		0.62
C08D	Selective calcium channel blockers with direct cardiac effect	0.55
C10A	Lipid modifying agents		0.71
M01A	Antiinflammatory and antirheumatic products, non-steroids					0.48
M05B	Drugs affecting bone structure and mineralisation			0.38	−0.26
N02A	Opioids					0.77
N02B	Other analgesics and antipyretics					0.62
N06D	Anti-dementia drugs				−0.86
R03A	Adrenergic inhalants	0.95
R03B	Other drugs for obstructive airway diseases, inhalants	0.95

^aDrugs for obstructive pulmonary diseases; ^bDrug for coronary diseases and hypertension; ^cDrugs for osteoporosis; ^dDrugs for heart failure; ^eDrugs for pain.

The table expresses the eigenvalue for each factor, their cumulative percent (proportion of variance of the drug data, explainable by the patterns) and the factorloading of the ATC 3rd level substances, whereby factorloadings less than 0.25 were omitted. All ATC 3rd level substances loading with a factorloading of 0.25 or more on one factor were included in one of the described drug patterns. Negative factorloadings ≥ 0.25 express a negative association between the drugs and the drug pattern described.

Table 3. Loading of factors with eigenvalue ≥ 1 and cumulative percent for ATC 3rd level substances of 1891 female patients in Germany (2008–2009).

Female
		D-Osteo^a	D-CDH^b	D-OPD^c	D-DG^d
	Eigenvalue	4.11	3.75	2.35	1.95
	Cumulative percent [%]	9.14	16.92	24.68	30.85
ATC 3rd level		Factorloading
A10A	Insulins and analogues		0.33
A10B	Blood glucose lowering drugs, excl. Insulins	−0.26
A11C	Vitamin A and D, including combinations of the two	0.96
A12A	Calcium	0.91
B01A	Antithrombotic agents		0.60
C01A	Cardiac glycosides		0.83
C01D	Vasodilators used in cardiac diseases		0.40
C03A	Low-ceiling diuretics – thiazides				0.61
C03C	High-ceiling diuretics		0.65
C03D	Potassium-sparing agents		0.27		0.65
M04A	Antigout preparations				0.75
M05B	Drugs affecting bone structure and mineralisation	0.84
N04B	Dopaminergic agents			−0.43
R03A	Adrenergic inhalants			0.94
R03B	Other drugs for obstructive airway diseases, inhalants			0.96
S01E	Antiglaucoma preparations and miotics				−0.31

^aDrugs for osteoporosis; ^bDrug for coronary diseases and hypertension; ^cDrugs for obstructive pulmonary diseases; ^dDiuretic drugs and drugs for gout.

The table expresses the eigenvalue for each factor, their cumulative percent (proportion of variance of the drug data, explainable by the patterns) and the factorloading of the ATC 3rd level substances, whereby factor loadings of less than 0.25 were omitted. All ATC 3rd level substances loading with a factorloading of 0.25 or more on one factor were included in one of the described drug patterns. Negative factorloadings ≥ 0.25 express a negative association between the drugs and the drug pattern described.

Figure 1. Overlapping of drug patterns (total number of patients) related to the total male population (1298) in Germany (2008–2009).

Figure 2. Overlapping of drug patterns (total number of patients) related to the total female population (1891) in Germany (2008–2009).

Table 4. Correlation coefficients ρ according to Spearman, p-values and 95% confidence interval of drug patterns and multimorbidity clusters for 3189 patients – separated for male and female – expressing the association between drug patterns and multimorbidity clusters from MultiCare study in Germany (2008–2009).

			95% confidence interval
	ρ	p-Value	Min	Max
Male
D-OPD^a x CMD^g	0.008	0.779	−0.049	0.063
D-OPD x ADS/P^h	0.088	0.002	0.032	0.145
D-OPD x NPS^i	−0.036	0.191	−0.047	−0.025
D-CDH^b x CMD	0.376	<0.001	0.322	0.430
D-CDH x ADS/P	−0.089	0.001	−0.144	−0.037
D-CDH x NPS	0.031	0.266	−0.021	0.073
D-Osteo^c x CMD	0.078	0.005	0.029	0.125
D-Osteo x ADS/P	0.019	0.486	−0.038	0.071
D-Osteo x NPS	0.022	0.434	−0.037	0.087
D-HF^d x CMD	0.089	0.001	0.052	0.121
D-HF x ADS/P	−0.011	−0.704	−0.061	0.042
D-HF x NPS	0.007	0.804	−0.028	0.068
D-Pain^e x CMD	−0.021	0.443	−0.081	0.037
D-Pain x ADS/P	0.103	<0.001	0.046	0.161
D-Pain x NPS	0.027	0.325	−0.033	0.101
Female
D-Osteo x CMD	−0.112	<0.001	−0.155	−0.065
D-Osteo x ADS/P	0.102	<0.001	0.060	0.142
D-Osteo x NPS	0.002	0.918	−0.041	0.050
D-CDH x CMD	0.301	<0.001	0.264	0.340
D-CDH x ADS/P	−0.103	<0.001	−0.150	−0.056
D-CDH x NPS	0.232	<0.001	0.178	0.284
D-OPD x CMD	−0.003	0.902	−0.048	0.040
D-OPD x ADS/P	0.094	<0.001	0.053	0.135
D-OPD x NPS	−0.180	0.436	−0.057	0.026
D-DG^f x CMD	0.157	<0.001	0.117	0.197
D-DG x ADS/P	0.0001	0.983	−0.045	0.044
D-DG x NPS	0.166	<0.001	0.056	0.172

^aDrugs for obstructive pulmonary diseases; ^bDrug for coronary diseases and hypertension, ^cDrugs for osteoporosis; ^dDrugs for heart failure; ^eDrugs for pain; ^fDiuretic drugs and drugs for gout; ^gCardiovascular and metabolic disorders; ^hAnxiety; depression; somatoform disorders and pain; ⁱNeuropsychiatric disorders.

Significant p-values are marked in bold.