Validity of Prescription-Defined and Hospital-Diagnosed Hypertension Compared with Self-Reported Hypertension in Denmark

Figures & data

Table 1 Characteristics of First-Time Responders of the Danish National Health Surveys (2010–2017) at the Time of Survey Response According to Self-Reported, Prescription-Defined, and Hospital-Diagnosed Hypertension

Characteristics	All Survey Responders N=442,490	Hypertension
		Self-Reported N=125,509	Prescription-Defined* N=108,040	Hospital-Diagnosed^† N=30,463
Female sex — N (%)	238,247 (53.8)	65,000 (51.8)	55,735 (51.6)	15,925 (52.3)
Age, years — Median (IQR)	53.1 (39.5–65.8)	65.0 (55.4–73.0)	67.6 (59.6–74.9)	69.1 (61.0–76.7)
Smoking — N (%)
Current	203,203 (45.9)	49,147 (39.2)	40,962 (37.9)	11,090 (36.4)
Former	140,297 (31.7)	50,388 (40.1)	45,985 (42.6)	13,537 (44.4)
Never	93,583 (21.1)	23,644 (18.8)	18,821 (17.4)	5073 (16.7)
Body mass index — N (%)
<18.5	8322 (1.9)	1424 (1.1)	1242 (1.1)	411 (1.3)
18.5–24.9	201,644 (45.6)	38,637 (30.8)	32,764 (30.3)	8785 (28.8)
25.0–29.9	152,895 (34.6)	50,039 (39.9)	43,249 (40.0)	11,759 (38.6)
≥30	67,417 (15.2)	30,963 (24.7)	26,684 (24.7)	8215 (27.0)
Physical activity — N (%)
Low	11,094 (2.5)	862 (0.7)	530 (0.5)	141 (0.5)
Moderate	281,272 (63.6)	74,475 (59.3)	63,153 (58.5)	16,012 (52.6)
High	133,584 (30.2)	43,169 (34.4)	37,594 (34.8)	12,098 (39.7)
Alcohol intake — N (%)
Low	335,587 (75.8)	91,221 (72.7)	78,319 (72.5)	22,393 (72.5)
Moderate or high	89,799 (20.3)	27,018 (21.5)	22,723 (21.0)	5815 (19.1)
Comorbidities — N (%)
Myocardial infarction	7903 (1.8)	5693 (4.5)	6839 (6.3)	2924 (9.6)
Congestive heart failure	6167 (1.4)	4250 (3.4)	5424 (5.0)	2538 (8.3)
Peripheral vascular disease	6995 (1.6)	4937 (3.9)	5213 (4.8)	2537 (8.3)
Cerebrovascular disease	13,418 (3.0)	8855 (7.1)	8746 (8.1)	4933 (16.2)
Dementia	1383 (0.3)	759 (0.6)	760 (0.7)	464 (1.5)
Chronic pulmonary disease	16,476 (3.7)	7117 (5.7)	6973 (6.5)	3366 (11.0)
Rheumatic disease	5191 (1.2)	2495 (2.0)	2498 (2.3)	1066 (3.5)
Peptic ulcer disease	3677 (0.8)	2048 (1.6)	2038 (1.9)	1021 (3.4)
Mild liver disease	2622 (0.6)	1171 (0.9)	1006 (0.9)	441 (1.4)
Diabetes, no chronic complications	14,855 (3.4)	11,056 (8.8)	11,105 (10.3)	6053 (19.9)
Diabetes, with chronic complications	4198 (0.9)	3339 (2.7)	3441 (3.2)	2068 (6.8)
Hemi- or paraplegia	343 (0.1)	127 (0.1)	115 (0.1)	59 (0.2)
Renal disease	2641 (0.6)	2102 (1.7)	2105 (1.9)	1327 (4.4)
Moderate or severe liver disease	459 (0.1)	234 (0.2)	226 (0.2)	98 (0.3)
Cancer	21,986 (5.0)	9991 (8.0)	9790 (9.1)	4113 (13.5)
Drug usage — N (%)
Antiplatelets	34,363 (7.8)	25,167 (20.1)	27,279 (25.2)	10,281 (33.7)
Anticoagulants	10,338 (2.3)	6807 (5.4)	8269 (7.7)	3607 (11.8)
ACE inhibitors	35,216 (8.0)	32,171 (25.6)	34,259 (31.7)	9262 (30.4)
A-II-R blockers	28,302 (6.4)	26,808 (21.4)	27,793 (25.7)	8180 (26.9)
Beta-blockers	30,915 (7.0)	24,918 (19.9)	30,147 (27.9)	9745 (32.0)
Calcium channel blockers	32,378 (7.3)	30,117 (24.0)	31,741 (29.4)	9869 (32.4)
Diuretics	53,646 (12.1)	47,109 (37.5)	50,472 (46.7)	14,541 (47.7)
Statins	50,772 (11.5)	36,006 (28.7)	37,521 (34.7)	12,041 (39.5)
Glucocorticoids	12,353 (2.8)	5240 (4.2)	5149 (4.8)	1755 (5.8)
Opioids	19,220 (4.3)	9641 (7.7)	9451 (8.7)	3816 (12.5)
Paracetamol	29,291 (6.6)	15,413 (12.3)	15,407 (14.3)	5947 (19.5)
Respiratory medications	22,982 (5.2)	9945 (7.9)	9841 (9.1)	3272 (10.7)
Proton pump inhibitors	29,786 (6.7)	15,491 (12.3)	15,552 (14.4)	5944 (19.5)
Antidepressants	26,150 (5.9)	10,776 (8.6)	10,140 (9.4)	3720 (12.2)
Antipsychotics	5093 (1.2)	1835 (1.5)	1697 (1.6)	622 (2.0)
NSAID	32,528 (7.4)	12,846 (10.2)	11,907 (11.0)	3352 (11.0)
Anti-diabetics	19,447 (4.4)	14,337 (11.4)	14,684 (13.6)	5472 (18.0)

Note: The columns may not add to 100% due to missing values. *Hypertension defined by the algorithm ≥1 prescription in ≥2 antihypertensive drug classes ≤365 days before survey response (see elaboration in Table 2). ^†Hypertension diagnosis within five years before survey response.

Abbreviations: ACE, angiotensin converting enzyme; A-II-R, angiotensin II receptor; NSAID, non-steroidal anti-inflammatory drugs.

Table 2 Validity of Algorithms for Prescription-Defined Hypertension Compared with Self-Reported Hypertension

Antihypertensive Drug Classes and Lookback Periods	Algorithm for Prescription-Defined Hypertension*
	≥2 Prescriptions in ≥2 drug Classes	≥2 Prescriptions in ≥1 drug Class	≥1 Prescription in ≥2 drug Classes	≥1 Prescription in ≥1 drug Class
	PPV/NPV; Sens/Spec	PPV/NPV; Sens/Spec	PPV/NPV; Sens/Spec	PPV/NPV; Sens/Spec
90-day lookback
Unspecified drugs^†	91/89; 5/100	92/80; 36/99	87/76; 22/99	84/88; 68/95
Specified drugs^‡	89/72; 3/100	94/78; 30/99	86/75; 15/99	87/88; 67/96
180-day lookback
Unspecified drugs^†	93/78; 29/99	93/79; 35/99	87/87; 63/96	87/88; 68/96
Specified drugs^‡	94/77; 24/99	93/80; 39/99	88/86; 59/97	86/91; 77/95
365-day lookback
Unspecified drugs^†	93/79; 34/99	93/80; 36/99	85/88; 67/95	84/88; 68/95
Specified drugs^‡	94/91; 30/99	93/80; 44/99	88/94; 75/96	85/92; 79/94

Notes: *The algorithms are based on the minimum number of filled prescriptions of the same drug (eg, enalapril or amlodipine) in the minimum number of different antihypertensive drug classes (eg, angiotensin-converting enzyme inhibitors or calcium channel blockers) within the specified lookback period. Eg, to have hypertension according to the ≥2 prescriptions in ≥2 drug classes algorithm when using a 90-day lookback, an individual could have ≥2 prescription fillings for enalapril and ≥2 prescription fillings for amlodipine within 90 days before self-reported hypertension. ^†Any drug within the antihypertensive drug classes: Angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, beta-blockers, calcium channel blockers, non-loop diuretics, alfa-blockers, and vasodilators. ^‡Antihypertensive specific drugs (ie, excluding heart failure specific drugs): Angiotensin-converting enzyme inhibitors (except trandolapril), angiotensin receptor blockers (except sacubitril/valsartan), beta-blockers (except nebivolol), calcium channel blockers, and thiazides.

Abbreviations: NPV, negative predictive value (%); PPV, positive predictive value (%); Sens, Sensitivity (%); Spec, Specificity (%).

Table 3 Validity of Hospital-Diagnosed Hypertension Compared with Self-Reported Hypertension

		Self-Reported Hypertension (Reference)
		Yes	No
Hospital-diagnosed hypertension	Yes	True positives (TP) N=27428	False positives (FP) N=3035	Positive predictive value = TP / (TP + FP) = 90 (90–90)
	No	False negatives (FN) N=98081	True negatives (TN) N=313946	Negative predictive value = TN / (FN + TN) = 76 (76–76)
		Sensitivity = TP / (TP + FN) = 22 (22–22)	Specificity = TN / (FP + TN) = 99 (99–99)

Note: The validity measures are presented in % (95% confidence interval).