Abstract
AIMS
Guidelines on the management of hypertension have recommended baseline testing of serum electrolyte and creatinine concentrations before treatment since the 1990s. We wished to examine the extent of laboratory monitoring in patients with newly diagnosed hypertension and newly treated with antihypertensive drugs.
METHODS
We carried out a retrospective analysis of 74 096 patients in the General Practice Research Database (GPRD) aged 18 years and older with newly diagnosed hypertension and prescribed a single antihypertensive agent. We determined the number of patients with a laboratory test for serum electrolyte and creatinine (or urea) concentrations prior to the first antihypertensive drug prescription and in the 6 months after and patient factors associated with subsequent monitoring.
RESULTS
Thirty-four thousand nine hundred and forty-seven patients (47%) had at least one biochemical test in the 12 months prior to beginning antihypertensive treatment, and 26 946 (36%) had at least one biochemical monitoring test in the 6 months after beginning antihypertensive treatment. Thirteen thousand five hundred and four (18%) had both baseline and monitoring tests. Baseline tests were normal in 11 671 patients (86%), of whom 10 213 (88%) had normal tests at first monitoring. Monitoring was significantly more likely in patients treated with ACE inhibitors than thiazides (adjusted OR 1.90; 95% CI 1.80, 2.00), older patients (adjusted OR 1.23; 95% CI 1.11, 1.36) [individuals aged 80–89 years compared with <40 years], and patients with diabetes mellitus (adjusted OR 2.03; 95% CI 1.91, 2.16).
CONCLUSION
Biochemical testing at baseline and monitoring after starting treatment is often omitted in newly diagnosed hypertensive patients. When both are tested, one in eight normal results becomes abnormal.
Keywords: adverse drug reactions, antihypertensive agents, drug monitoring
WHAT IS ALREADY KNOWN ABOUT THIS SUBJECT
Guidelines recommend biochemical monitoring of patients treated with antihypertensive agents, although there is little primary evidence for these recommendations.
WHAT THIS STUDY ADDS
Patients treated for hypertension in general practice often have no biochemical tests before, or in the 6 months after, starting drug treatment.
Introduction
Monitoring is a process of checking a system that changes with time, in order to guide changes to the system that will maintain it or improve it [1]. Monitoring of drug treatment can have several effects: improved adherence, better selection of drug therapy and better titration of treatment [2]. Monitoring can also, perhaps most importantly, identify potential adverse reactions to treatment.
Guidelines recommend that patients with newly diagnosed hypertension should have baseline biochemical tests of renal function and electrolyte concentrations before treatment, monitored again 1 week and 1 month after starting treatment, and at intervals following any dose changes [3–5]. These recommendations are based on the supposition that pre-treatment testing can discover rare secondary causes of hypertension and identify contraindications to treatment, and that post-treatment biochemical monitoring can identify changes related to adverse drug reactions (ADRs) before they have caused serious or permanent effects, and so avert them [1, 6]. However, primary evidence for the recommendations is lacking. A questionnaire survey suggested that general practitioners (GPs) often neglect biochemical monitoring [7].
We examined the extent and effect of laboratory monitoring in a cohort of patients in UK general practice with newly diagnosed hypertension and newly treated with antihypertensive drugs.
Methods
The study protocol was approved by the Independent Scientific Advisory Committee for Medicines and Healthcare products Regulatory Agency (MHRA) database research.
Data source
The GPRD is an automated medical record database that contains anonymized longitudinal records from over 460 general practices with approximately 3.5 million patients in the UK [8]. The computerized information includes demographic data, clinical history, laboratory tests and drug prescriptions. The age and sex distributions of patients in the database are representative of those of the UK general population [9], and its completeness and validity are well documented [10, 11].
Study population
We derived the study cohort from the full GPRD dataset. Selection into the cohort required patients to be 18 years or older and to have a Read or OXMIS (Oxford Medical Information System) code compatible with a diagnosis of essential hypertension, or three blood pressure measurements of greater than 160/100 mmHg (the threshold for initiation of antihypertensive treatment without other cardiovascular risk factors at the time) on or within 365 days prior to the first prescription for a single antihypertensive drug (index date), or both. We then identified all patients in the cohort by selecting patients with a first prescription between 1 January 2000 and 31 December 2003 for a single antihypertensive drug from seven defined drug classes: angiotensin-converting enzyme inhibitors (ACE inhibitor), angiotensin-II (AT-II) receptor antagonists, calcium-channel blockers, thiazide diuretics, α-adrenoceptor blocking drugs (α-blockers), β-adrenoceptor blocking drugs (β-adrenoceptor antagonists), and loop or aldosterone antagonist/potassium-sparing diuretics. We defined this as a first prescription of antihypertensive drug therapy provided there had been no antihypertensive drug therapy in the preceding year.
Using methods described previously [12], we identified 1148 women who could have been pregnant between 1 January 2000 and 31 December 2003 and excluded them from the analysis because of potential differences in their clinical course, treatment and monitoring.
Assessment of covariates
Smoking status and body mass index (BMI) were used if recorded within 5 years of the index date and systolic blood pressure and diastolic blood pressure were used if recorded in the year prior to the index date. We used the British Hypertension Society guidelines to classify the severity of hypertension [13]. Socio-economic status was inferred from the social deprivation scores assigned to the postcode of each general practice using the Index of Multiple Deprivation.
Assessment of laboratory measurements
We defined baseline biochemical testing as a measurement of one or more of serum creatinine, urea, potassium or sodium concentrations between 365 days prior to the index date and 2 days following it, to allow for any delays in the uploading of laboratory test results to the general practice (prior to 2003 the majority of laboratory data entry was not automated). Follow-up monitoring was defined as a measurement of one or more of the serum concentrations between 3 days and 6 months after the index date. Where both urea and creatinine concentration were measured, we used only the latter in the analyses.
Assessment of abnormal values
Abnormal laboratory values at baseline and first follow-up monitoring test were determined using standardized reference ranges provided with the tests. Standardized reference ranges were 56–122 µmol·l−1 for creatinine, 2.6–7.2 mmol·l−1 for urea, 135–146 mmol·l−1 for sodium, and 3.5–5.1 mmol·l−1 for potassium. Where such ranges were missing, they were inferred from the patient's gender and age. Values outside the reference ranges for sodium or potassium concentrations, and above the reference ranges for creatinine or urea, were classified as abnormal.
Statistical analysis
We used Microsoft Access® to arrange data, and Stata® 10.0 to perform statistical analyses. We determined the relationships between patient variables and the probability of any monitoring within 6 months in univariable analyses, and then entered baseline characteristics that were statistically significant at the P < 0.05 level or were biologically plausible into a multivariable logistic regression model using backwards stepwise variable selection. Gender was forced into the model a priori.
Results
We identified 74 096 patients, 48% male, who were first prescribed a single antihypertensive agent between January 2000 and December 2003 (Table 1). Patients most often began treatment with a thiazide diuretic (48.1%), followed by β-adrenoceptor antagonists (24.3%), and ACE inhibitors (15.2%). A greater proportion of patients aged below 40 years was treated with an ACE inhibitor. Thirty-four thousand nine hundred and forty-seven patients (47.2%) had one or more laboratory tests in the year prior to the index date. The majority of baseline testing (57%) occurred 1 month before the start of treatment, and the modal time interval was 0 (the day of the initiation of antihypertensive therapy). Patients with baseline testing were more likely to be male, younger, have diabetes mellitus or smoke tobacco.
Table 1.
Baseline patient characteristics
| Characteristic | Total (n = 74 096) | Any baseline testing (n = 34 947) | No baseline testing (n = 39 149) | P value |
|---|---|---|---|---|
| Male, n (%) | 35 345 (47.7) | 17 394 (49.8) | 17 951 (45.9) | <0.0005 |
| Mean age (years) (SD) | 61.1 (12.8) | 60.9 (12.5) | 61.2 (13.1) | 0.024 |
| BMI (kg m−2) | ||||
| Mean (SD) | 28.5 (5.4) | 28.5 (5.4) | 28.4 (5.4) | 0.0002 |
| Underweight, n (%) | 654 (0.9) | 312 (0.9) | 342 (0.9) | <0.0005† |
| Normal, n (%) | 16 039 (21.6) | 7 485 (21.4) | 8 554 (21.8) | |
| Overweight, n (%) | 25 392 (34.3) | 12 426 (35.6) | 12 966 (33.1) | |
| Obese, n (%) | 20 606 (27.8) | 10 115 (28.9) | 10 491 (26.7) | |
| Smoking, n (%) | ||||
| Ever | 33 118 (44.7) | 16 256 (46.5) | 16 862 (43.1) | <0.0005 |
| Never | 38 166 (51.5) | 17 679 (50.6) | 20 487 (52.3) | |
| Presence of diabetes mellitus, n (%) | 6 437 (8.7) | 4 565 (13.1) | 1 872 (4.8) | <0.0005 |
| SES quintile, n (%) | ||||
| 0 (least deprived) | 15 660 (21.1) | 6 892 (19.7) | 8 768 (22.4) | <0.0005† |
| 1 | 13 205 (17.8) | 6 575 (18.8) | 6 630 (16.9) | |
| 2 | 15 005 (20.3) | 6 886 (19.7) | 8 119 (20.7) | |
| 3 | 14 154 (19.1) | 6 034 (17.3) | 8 120 (20.7) | |
| 4 (most deprived) | 16 072 (21.7) | 8 560 (24.5) | 7 512 (19.2) | |
| Blood pressure (mmHg) (SD) | ||||
| Mean systolic blood pressure | 170.9 (20.1) | 170.5 (19.5) | 171.3 (20.6) | <0.0005 |
| Mean diastolic blood pressure | 97.2 (11.3) | 97.0 (11.3) | 97.5 (11.4) | <0.0005 |
| Drug class initiated, n (%) | ||||
| ACE inhibitor | 11 245 (15.2) | 6 609 (18.9) | 4 636 (11.8) | <0.0005 |
| α- blocker | 753 (1.0) | 342 (1.0) | 411 (1.1) | 0.335 |
| AT-II receptor antagonist | 1 630 (2.2) | 769 (2.2) | 861 (2.2) | 0.991 |
| β-adrenoceptor antagonist | 17 977 (24.3) | 7 899 (22.6) | 10 078 (25.7) | <0.0005 |
| Calcium channel blocker | 5 651 (7.6) | 2 574 (7.4) | 3 077 (7.9) | 0.011 |
| Loop or K-sparing diuretic | 416 (0.6) | 181 (0.5) | 235 (0.6) | 0.134 |
| Thiazide diuretic | 35 653 (48.1) | 16 302 (46.7) | 19 351 (49.4) | <0.0005 |
| Combination preparation | 771 (1.0) | 271 (0.8) | 500 (1.3) | <0.0005 |
P values presented are from the t-test for continuous variables and the Chi-squared test for binary variables. P values from the Chi-squared test for trend are indicated with a.
Some data for BMI and smoking status were missing.
In the 6 months after the date of first prescription 26 946 (36.4%) patients had at least one biochemical measurement. Of those patients who were monitored at least once within 6 months, 21 194 (78.7%) had all three tests of potassium, sodium and renal function and 1557 (5.8%) of those who were monitored had only one test. In patients with at least one follow-up test, 27% of patients had more than one follow-up monitoring test (Table 2). The mean number of repeat monitoring tests in those who had at least one test was 1.4 (Table 3). Patients started with an ACE inhibitor or a potassium-sparing diuretic had a greater mean number of follow-up monitoring tests. In 13 504/74 096 patients (18.2%), both baseline testing and follow-up monitoring were performed.
Table 2.
Summary for serum creatinine, concentration (micromoles per litre) and urea, sodium, and potassium concentrations (millimoles per litre) measured at baseline and in the 6 months after the initiation of treatment
| Creatinine | Urea | Urea (when creatinine missing) | Sodium | Potassium | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Baseline testing | Monitoring 6 months after index date | Baseline testing | Monitoring 6 months after index date | Baseline testing | Monitoring 6 months after index date | Baseline testing | Monitoring 6 months after index date | Baseline testing | Monitoring 6 months after index date | |
| Number of patients with at least one test, n (%) | 33 976 (45.9) | 26 282 (35.4) | 28 331 (38.2) | 22 456 (30.3) | 794 (1.1) | 524 (0.7) | 31 615 (42.7) | 24 853 (33.5) | 31 404 (42.4) | 24 804 (33.5) |
| Mean value | 89.2 | 91.3 | 5.1 | 5.5 | 5.6 | 6.1 | 140.1 | 139.5 | 4.4 | 4.3 |
| Median value | 87.0 | 89.0 | 5.0 | 5.3 | 5.4 | 5.8 | 140.0 | 140.0 | 4.4 | 4.3 |
| Range of values | 23–905 | 25–1107 | 0.3–68 | 0.2–70 | 2.1–20 | 1.7–70 | 117–174 | 116–152 | 2.4–11.8 | 2.2–9.9 |
| Number of patients who were monitored with | ||||||||||
| one follow-up test, n (%) | – | 19 146 (72.8) | – | 16 441 (73.2) | – | 469 (89.5) | – | 18 116 (72.9) | – | 18 154 (73.2) |
| two follow-up tests, n (%) | – | 5 093 (19.4) | – | 4 314 (19.2) | – | 46 (8.8) | – | 4 825 (19.4) | – | 4 759 (19.2) |
| three follow-up tests, n (%) | – | 1 384 (5.3) | – | 1 157 (5.2) | – | 5 (1.0) | – | 1 296 (5.2) | – | 1 293 (5.2) |
| four follow-up tests, n (%) | – | 428 (1.6) | – | 360 (1.6) | – | 1 (0.2) | – | 405 (1.6) | – | 390 (1.6) |
| ≥five follow-up tests, n (%) | – | 231 (0.9) | – | 184 (0.8) | – | 3 (0.6) | – | 211 (0.8) | – | 208 (0.8) |
| Abnormal test, n (%) | 1 369 (4.0) | 1 587 (6.0) | 2 913 (10.3) | 3 495 (15.6) | 93 (11.7) | 98 (18.7) | 1 096 (3.5) | 1 483 (6.0) | 1 839 (5.9) | 2 058 (8.3) |
Table 3.
Number of laboratory tests in the 6 months after first prescription by drug class in those patients who had at least one test (n, number of patients with at least one test)
| Creatinine | Urea | Sodium | Potassium | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| n | Mean(median) | Range | n | Mean(median) | Range | n | Mean(median) | Range | n | Mean(median) | Range | |
| ACE inhibitor | 5 962 | 1.56 (1) | 1–9 | 4 887 | 1.56 (1) | 1–9 | 5 542 | 1.55 (1) | 1–9 | 5 540 | 1.55 (1) | 1–9 |
| α-blocker | 213 | 1.37 (1) | 1–8 | 172 | 1.36 (1) | 1–6 | 196 | 1.33 (1) | 1–5 | 195 | 1.32 (1) | 1–5 |
| AT-II receptor antagonist | 564 | 1.31 (1) | 1–7 | 474 | 1.29 (1) | 1–5 | 528 | 1.31 (1) | 1–7 | 525 | 1.31 (1) | 1–7 |
| β-adrenoceptor antagonist | 5 298 | 1.31 (1) | 1–15 | 4 526 | 1.30 (1) | 1–15 | 5 071 | 1.31 (1) | 1–15 | 5 055 | 1.31 (1) | 1–15 |
| Calcium channel blocker | 1 734 | 1.38 (1) | 1–12 | 1 500 | 1.38 (1) | 1–12 | 1 625 | 1.37 (1) | 1–12 | 1 631 | 1.37 (1) | 1–12 |
| Loop or K-sparing diuretic | 146 | 1.62 (1) | 1–8 | 115 | 1.44 (1) | 1–8 | 133 | 1.56 (1) | 1–8 | 129 | 1.57 (1) | 1–8 |
| Thiazide diuretic | 12 167 | 1.35 (1) | 1–14 | 10 617 | 1.34 (1) | 1–9 | 11 566 | 1.35 (1) | 1–14 | 11 541 | 1.35 (1) | 1–14 |
| Combination preparation | 198 | 1.29 (1) | 1–8 | 165 | 1.30 (1) | 1–8 | 192 | 1.29 (1) | 1–8 | 188 | 1.30 (1) | 1–8 |
| Total | 26 282 | 1.40 (1) | 1–15 | 22 456 | 1.38 (1) | 1–15 | 24 853 | 1.39 (1) | 1–15 | 24 804 | 1.38 (1) | 1–15 |
The median interval between first prescription and first biochemical monitoring was 35 days, and the modal interval was 7 days. The median interval between first prescription and first subsequent GP consultation was 23 days, and the modal interval was 28 days (Figure 1).
Figure 1.
Time to first monitoring and first GP consultation in the 6 months after the first antihypertensive prescription (index date). GP consultation (
); any biochemical monitoring(
)
One or more test results were abnormal for 4 078/34 947 (11.7%) patients at baseline, and 4 775/26 946 (17.7%) patients at first monitoring. The distribution of abnormal tests before and after the initiation of antihypertensive treatment is presented in Figure 2. The proportion of patients with an abnormal creatinine value (4% at baseline to 6% at first follow-up monitoring) and with an abnormal urea value (10% at baseline to 16% at first monitoring) increased (Table 2). Similar increases in the proportion of patients with abnormal tests at baseline and first follow-up monitoring were seen with the sodium (4% to 6%) and potassium tests (6% to 8%). In the subset of 13 504 patients who had both baseline tests and at least one monitoring test, the proportion of abnormal results increased from 14% to 18% with treatment. Of the 11 671 patients with normal results at baseline, 1 458 (12.5%) developed abnormal results at first monitoring. Only 14 patients had a recorded baseline potassium value of less than 3 mmol·l−1, of whom 8 (57%) had a follow-up serum potassium test. Eighty-five patients had a baseline creatinine value of greater than 200 µmol l−1, of whom 50 (59%) had a subsequent creatinine measurement.
Figure 2.
Distribution of any baseline test or any follow-up monitoring test in the 4 weeks before and after the first antihypertensive prescription (index date). Total number of tests (
); number of abnormal tests (
)
The percentage of patients who had biochemical follow-up monitoring in the first 6 months of treatment increased with time, from 25.6% in patients whose first prescription was in 2000 to 46.7% in patients whose first prescription was in 2003.
Logistic regression model
The patient factors associated with monitoring are shown in Table 4. Older patients, those with pre-existing diabetes, patients started on ACE inhibitors at baseline, and those with a first antihypertensive prescription in 2003 were more likely to be monitored at least once after starting antihypertensive treatment. In the multivariable model, monitoring was significantly associated with increasing age, the presence of diabetes, first-line antihypertensive therapy with ACE inhibitors and year of first prescription. Monitoring was also significantly associated with baseline testing and socio-economic status, being least likely for those of intermediate status.
Table 4.
Logistic regression model of any monitoring in 6 months following first prescription (n is the number of patients who exhibited the characteristic from N, the total population assessed for this characteristic)
| Characteristic | n/N | Unadjusted OR95% CI | P-value | Adjusted OR*95% CI | P value | ||
|---|---|---|---|---|---|---|---|
| Gender | |||||||
| Female | 13 768/38 751 | 1 | 1 | ||||
| Male | 13 178/35 345 | 1.08 | 1.05, 1.11 | <0.0005 | 1.00 | 0.97, 1.03 | 0.955 |
| Age (years) | |||||||
| <40 | 1 122/3 492 | 1 | 1 | ||||
| 40–49 | 3 599/10 388 | 1.12 | 1.03, 1.22 | 0.007 | 1.08 | 0.99, 1.18 | 0.086 |
| 50–59 | 7 118/20 360 | 1.14 | 1.05, 1.23 | 0.001 | 1.13 | 1.04, 1.23 | 0.004 |
| 60–69 | 7 214/19 312 | 1.26 | 1.17, 1.36 | <0.0005 | 1.22 | 1.12, 1.33 | <0.0005 |
| 70–79 | 5 830/14 971 | 1.35 | 1.25, 1.46 | <0.0005 | 1.33 | 1.22, 1.44 | <0.0005 |
| 80–89 | 1 936/5 190 | 1.26 | 1.15, 1.38 | <0.0005 | 1.23 | 1.11, 1.36 | <0.0005 |
| 90–100 | 127/383 | 1.04 | 0.84, 1.31 | 0.683 | 1.19 | 0.92, 1.53 | 0.191 |
| Smoking status | |||||||
| Never | 13 710/38 166 | 1 | 1 | ||||
| Ever | 12 537/33 118 | 1.09 | 1.05, 1.12 | <0.0005 | 1.06 | 1.03, 1.10 | <0.0005 |
| Diabetes status | |||||||
| No | 23 271/67 659 | 1 | 1 | ||||
| Yes | 3 675/6 437 | 2.54 | 2.41, 2.67 | <0.0005 | 2.03 | 1.91, 2.16 | <0.0005 |
| SES quintile | |||||||
| 0 (least deprived) | 5 823/15 660 | 1 | 1 | ||||
| 1 | 5 011/13 205 | 1.03 | 0.98, 1.08 | 0.182 | 1.03 | 0.98, 1.09 | 0.206 |
| 2 | 5 318/15 005 | 0.93 | 0.89, 0.97 | 0.002 | 0.90 | 0.85, 0.94 | <0.0005 |
| 3 | 4 760/14 154 | 0.86 | 0.82, 0.90 | <0.0005 | 0.86 | 0.82, 0.91 | <0.0005 |
| 4 | 6 034/16 072 | 1.02 | 0.97, 1.06 | 0.508 | 1.03 | 0.98, 1.08 | 0.232 |
| Hypertension | |||||||
| Normal or mild | 4 147/11 373 | 1 | 1 | ||||
| Moderate | 10 836/30 241 | 0.97 | 0.93, 1.02 | 0.232 | 1.09 | 1.03, 1.14 | 0.001 |
| Severe | 11 098/28 529 | 1.27 | 1.13, 1.45 | <0.0005 | 1.34 | 1.27, 1.40 | <0.0005 |
| BMI | |||||||
| Underweight | 243/654 | 1.01 | 0.86, 1.19 | 0.860 | – | ||
| Normal | 5 905/16 039 | 1 | – | ||||
| Overweight | 9 334/25 392 | 1.00 | 0.96, 1.04 | 0.907 | – | ||
| Obese | 7 840/20 606 | 1.05 | 1.00, 1.10 | 0.016 | – | ||
| Drug therapy | |||||||
| Thiazide diuretic | 12 492/35 653 | 1 | 1 | ||||
| ACE inhibitor | 6 049/11 245 | 2.16 | 2.07, 2.25 | <0.0005 | 1.90 | 1.80, 2.00 | <0.0005 |
| α-blocker | 220/753 | 0.71 | 0.65, 0.90 | 0.001 | 0.76 | 0.64, 0.90 | <0.0005 |
| AT-II receptor antagonist | 576/1 630 | 1.01 | 0.91, 1.12 | 0.804 | 0.97 | 0.87, 1.09 | 0.641 |
| β-adrenoceptor antagonist | 5 441/17 977 | 0.80 | 0.77, 0.84 | <0.0005 | 0.86 | 0.82, 0.90 | <0.0005 |
| Calcium channel blocker | 1 808/5 651 | 0.87 | 0.82, 0.93 | <0.0005 | 0.89 | 0.83, 0.94 | <0.0005 |
| Loop or K-sparing diuretic | 149/416 | 1.03 | 0.85, 1.27 | 0.740 | 1.29 | 1.03, 1.62 | 0.029 |
| Combination preparation | 211/771 | 0.70 | 0.60, 0.82 | <0.0005 | 0.73 | 0.62, 0.87 | <0.0005 |
| Any baseline testing | |||||||
| No | 13 442/39 149 | 1 | |||||
| Yes | 13 504/34 947 | 1.20 | 1.17, 1.24 | <0.0005 | 0.91 | 0.88, 0.94 | <0.0005 |
| Year of first antihypertensive prescription | |||||||
| 2000 | 4 251/16 508 | 1 | 1 | ||||
| 2001 | 5 742/17 927 | 1.36 | 1.30, 1.44 | <0.0005 | 1.33 | 1.27, 1.40 | <0.0005 |
| 2002 | 7 543/19 532 | 1.81 | 1.73, 1.90 | <0.0005 | 1.76 | 1.68, 1.85 | <0.0005 |
| 2003 | 9 410/20 129 | 2.53 | 2.42, 2.65 | <0.0005 | 2.51 | 2.39, 2.63 | <0.0005 |
Adjusted for gender, age, smoking status, diabetes, socio-economic status (SES), hypertension, drug therapy, any baseline testing and year of first antihypertensive prescription.
Discussion
In a large cohort of patients with newly diagnosed hypertension, most patients began treatment with either thiazide diuretics or β-adrenoceptor antagonists, consistent with contemporaneous guidelines [14] and as previously demonstrated in the UK [15]. Only 34 947 patients (47%) had any baseline biochemical test in the 12 months prior to beginning antihypertensive treatment. A minority (36%) of our cohort had follow-up laboratory monitoring in the 6 months after the initiation of drug therapy. This compares with similar poor rates of follow-up reported in other studies of antihypertensive drugs [16–19] and in other drugs such as statins [20, 21], thiazolidinediones [22] and allopurinol [23]. Half the cohort had a GP consultation within 24 days of starting treatment, so the lack of monitoring was likely not to be due to non-attendance but because only a small proportion of patients was monitored. We found older patients and those with diabetes mellitus were more likely to have either baseline testing or follow-up monitoring, which is consistent with the results of other smaller studies [16, 19].
The use of the GPRD allowed for the examination of real-life data on monitoring by UK GPs. Our work is, we believe, the first to look at monitoring in a range of antihypertensive medications in UK general practice. Although we have shown some similarities to previous studies, the results from prior work is of limited relevance to UK primary care as such research has been restricted to specific drug classes or to data from health maintenance organizations, where monitoring is often protocol-driven.
We used standardized reference ranges to define abnormality and identified that as many as one in eight patients will have an abnormal baseline test, although many of these abnormalities may be clinically insignificant. Given that 5% of the normal population will have a result outside the reference range, and 2.5% will have results above normal, the expected rate of abnormal creatinine, potassium or sodium results as we defined them was [1 − (0.975 × 0.95 × 0.95) = 0.12], that is, 12%. This implies that the observed rate of abnormal results at baseline is no different in antihypertensive-treated patients from normal subjects.
According to a questionnaire survey, a minority of European GPs (17%) routinely request the laboratory tests recommended to assess hypertensive patients [24]. Absence of baseline testing is inconsistent with standard guidelines. Hypertension guidelines have, for more than 15 years, stated that the measurement of serum electrolyte and urea or creatinine concentrations is ‘essential’ [25]. Simple baseline investigations allow the detection of some causes of secondary hypertension, associated cardiovascular risk factors, evidence of target organ damage, and co-morbid diseases, all of which can influence treatment decisions. Absence of baseline testing also makes changes in renal function and electrolyte concentrations more difficult to assess.
It has been recommended that all hypertensive patients given an ACE inhibitor should have their renal function measured before and after 1 week of treatment [4]. This is because a rise in previously normal serum creatinine concentration of ≥30% after starting treatment with an ACE inhibitor may suggest renovascular disease. Few patients had both baseline testing and subsequent monitoring, depriving their doctors of the opportunity to assess intra-individual changes within the reference range. Even for those patients treated with an ACE inhibitor, only 38% had both baseline testing and any follow-up monitoring. In patients who did have at least one follow-up monitoring test, we found that less than a third of patients did not have a second follow-up test. This limits the ability of GPs to judge sequential changes in serum concentrations.
A trade-off exists between monitoring too frequently and not enough. Too frequent monitoring may identify short-term changes in renal function, which may lead to the inappropriate cessation of antihypertensive therapy to the detriment of blood pressure control. Too few monitoring tests would fail to identify significant adverse effects of treatment.
The proportion of patients monitored after starting treatment increased over the 4 year period that we examined. This improvement was similar to the pattern demonstrated in the proportion of patients with a record for weight and smoking, and may reflect a general improvement in standards in general practice [26], or in the electronic recording of test results, or in the way guidelines are implemented.
Although patients treated with diuretic drugs and those acting on the angiotensin system were more likely to be monitored than those treated with drugs such as calcium antagonists that have fewer biochemical adverse effects, many patients treated with drugs with a potential to cause biochemical harm went unmonitored. The odds of being monitored were not significantly influenced by gender but were increased in older patients and those with diabetes, even when adjustment was made for other factors.
Limitations
Coding for hypertension may not have been consistent across general practices. We may also have missed patients treated with antihypertensive agents at blood pressures lower than we defined. We relied on records of issued prescriptions, but we did not know whether the patient collected, or took, the medicine. We might therefore have over-estimated the number of patients taking medicines, and so over-estimated the ‘non-monitoring’ rate [27].
The general practice records may not have captured all the monitoring in secondary care. However, most patients with hypertension, especially those with ‘simple’ hypertension requiring therapy with a single agent, are treated in the community. The GP who initiated treatment would usually be responsible for monitoring.
We did not include renal function at baseline as a potential predictor of subsequent monitoring in our models. That would have required a measurement of serum creatinine or urea concentration and biased the results by restricting the analyses to patients who already had a laboratory test and might be more likely to be monitored.
Our data were collected during a time of change in United Kingdom general practice that resulted in improvements in the management of hypertension as judged by the increased recording of blood pressure measurements [28]. The proportion who had at least one monitoring test increased during the survey period. Although we cannot be sure our data reflect current practice in biochemical monitoring, the differences in monitoring between drug class and among patients are likely to persist.
Finally, we considered results abnormal if they fell outside the reference range for serum sodium or potassium concentrations or above the upper limit of the reference range for serum creatinine or urea concentrations. In practice, clinicians do not act on every abnormal result, and the threshold values for action will often be some way outside the defined reference range. Therefore, there will be fewer patients with results sufficiently abnormal to warrant clinical intervention than with abnormal results as we defined them. We were able to allow for this to some extent in estimating renal function. Serum urea concentration was often outside the reference range, but where both urea and creatinine concentrations were measured, we used only the latter.
Guidelines recommend baseline and post-treatment biochemical testing of patients with hypertension. Judged by these standards, current practice represents an oversight (‘an accidental omission; a mistake made through inadvertence or negligence’) rather than oversight (‘the action of overseeing something; supervision, inspection’) [29]. The elderly, those with diabetes, and those on ACE inhibitors were more likely to be monitored. Few patients had both baseline tests and subsequent monitoring, limiting the ability of GPs to assess changes in serum biochemical measurements for each patient. Future studies are necessary in order to determine whether more frequent or more assiduous biochemical monitoring in hypertensive patients in general practice would reduce harm from adverse drug reactions. Results from such studies would help to devise rational monitoring schemes.
Acknowledgments
This study is based in part on data from the Full Feature General Practice Research Database obtained under licence from the UK Medicines and Healthcare products Regulatory Agency. However, the interpretation and conclusions contained in this study are those of the author/s alone.
We thank Mr Christopher Anton for his assistance with the data manipulation and extraction.
Competing interests
There are no competing interests to declare.
Access to the General Practice Research Database (GPRD) database was funded through the Medical Research Council's licence agreement with the MHRA. Sarah E. McDowell was supported by the Antidote Trust Fund of Sandwell and West Birmingham Hospitals NHS Trust.
REFERENCES
- 1.Coleman JJ, Ferner RE, Evans SJ. Monitoring for adverse drug reactions. Br J Clin Pharmacol. 2006;61:371–8. doi: 10.1111/j.1365-2125.2006.02596.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Glasziou P, Irwig L, Mant D. Monitoring in chronic disease: a rational approach. BMJ. 2005;330:644–8. doi: 10.1136/bmj.330.7492.644. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Martin U, Coleman JJ. Monitoring renal function in hypertension. BMJ. 2006;333:896–9. doi: 10.1136/bmj.38987.502512.DE. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Smellie WS, Forth J, Coleman JJ, Irvine W, Dore PC, Handley G, Williams DG, Galloway PJ, Kerr KG, Herriot R, Spickett GP, Reynolds TM. Best practice in primary care pathology: review 6. J Clin Pathol. 2007;60:225–34. doi: 10.1136/jcp.2006.040014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Hippisley-Cox J, Yeo W, Ramsay L. Monitoring requirements for cardiovascular drugs. Prescriber. 2000;11:43–56. [Google Scholar]
- 6.Pirmohamed M, Ferner RE. Monitoring drug treatment. BMJ. 2003;327:1179–81. doi: 10.1136/bmj.327.7425.1179. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kalra PA, Kumwenda M, MacDowall P, Roland MO. Questionnaire study and audit of use of angiotensin converting enzyme inhibitors and monitoring in general practice: the need for guidelines to prevent renal failure. BMJ. 1999;318:234–7. doi: 10.1136/bmj.318.7178.234. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.General Practice Research Database. GPRD – An Overview. London: Crown Copyright; 2007. [Google Scholar]
- 9.Gelfand JM, Margolis DJ, Dattani H. The UK General Practice Research Database. In: Strom BL, editor. Pharmacoepidemiology. 4th. Chichester: John Wiley & Sons Ltd.; 2005. pp. 337–46. [Google Scholar]
- 10.Jick SS, Kaye JA, Vasilakis-Scaramozza C, Garcia Rodriguez LA, Ruigomez A, Meier CR, Schlienger RG, Black C, Jick H. Validity of the general practice research database. Pharmacotherapy. 2003;23:686–9. doi: 10.1592/phco.23.5.686.32205. [DOI] [PubMed] [Google Scholar]
- 11.Lewis JD, Brensinger C, Bilker WB, Strom BL. Validity and completeness of the General Practice Research Database for studies of inflammatory bowel disease. Pharmacoepidemiol Drug Saf. 2002;11:211–8. doi: 10.1002/pds.698. [DOI] [PubMed] [Google Scholar]
- 12.Hardy JR, Holford TR, Hall GC, Bracken MB. Strategies for identifying pregnancies in the automated medical records of the General Practice Research Database. Pharmacoepidemiol Drug Saf. 2004;13:749–59. doi: 10.1002/pds.935. [DOI] [PubMed] [Google Scholar]
- 13.Williams B, Poulter NR, Brown MJ, Davis M, McInnes GT, Potter JF, Sever PS, Thom SM. British Hypertension Society guidelines for hypertension management 2004 (BHS-IV): summary. BMJ. 2004;328:634–40. doi: 10.1136/bmj.328.7440.634. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Ramsay L, Williams B, Johnston G, MacGregor G, Poston L, Potter J, Poulter N, Russell G. Guidelines for management of hypertension: report of the third working party of the British Hypertension Society. J Hum Hypertens. 1999;13:569–92. doi: 10.1038/sj.jhh.1000917. [DOI] [PubMed] [Google Scholar]
- 15.Walley T, Duggan AK, Haycox AR, Niziol CJ. Treatment for newly diagnosed hypertension: patterns of prescribing and antihypertensive effectiveness in the UK. J R Soc Med. 2003;96:525–31. doi: 10.1258/jrsm.96.11.525. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Hurley JS, Roberts M, Solberg LI, Gunter MJ, Nelson WW, Young L, Frost FJ. Laboratory safety monitoring of chronic medications in ambulatory care settings. J Gen Intern Med. 2005;20:331–3. doi: 10.1111/j.1525-1497.2005.40182.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Clayton JA, Rodgers S, Blakey J, Avery A, Hall IP. Thiazide diuretic prescription and electrolyte abnormalities in primary care. Br J Clin Pharmacol. 2006;61:87–95. doi: 10.1111/j.1365-2125.2005.02531.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.McAlister FA, Tu K, Majumdar SR, Padwal R, Chen Z, Campbell NR. Laboratory testing in newly treated elderly hypertensive patients without co-morbidities: a population-based cohort study. Open Med. 2007;1:E60–7. [PMC free article] [PubMed] [Google Scholar]
- 19.Raebel MA, McClure DL, Simon SR, Chan KA, Feldstein A, Andrade SE, Lafata JE, Roblin D, Davis RL, Gunter MJ, Platt R. Laboratory monitoring of potassium and creatinine in ambulatory patients receiving angiotensin converting enzyme inhibitors and angiotensin receptor blockers. Pharmacoepidemiol Drug Saf. 2007;16:55–64. doi: 10.1002/pds.1217. [DOI] [PubMed] [Google Scholar]
- 20.Abookire SA, Karson AS, Fiskio J, Bates DW. Use and monitoring of ‘statin’ lipid-lowering drugs compared with guidelines. Arch Intern Med. 2001;161:53–8. doi: 10.1001/archinte.161.1.53. [DOI] [PubMed] [Google Scholar]
- 21.Tragni E, Filippi A, Mazzaglia G, Sessa E, Cricelli C, Catapano AL. Monitoring statin safety in primary care. Pharmacoepidemiol Drug Saf. 2007;16:652–7. doi: 10.1002/pds.1361. [DOI] [PubMed] [Google Scholar]
- 22.Graham DJ, Drinkard CR, Shatin D, Tsong Y, Burgess MJ. Liver enzyme monitoring in patients treated with troglitazone. JAMA. 2001;286:831–3. doi: 10.1001/jama.286.7.831. [DOI] [PubMed] [Google Scholar]
- 23.Raebel MA, McClure DL, Simon SR, Chan KA, Feldstein AC, Gunter MJ, Lafata JE, Harrold L, Nelson WW, Platt R. Frequency of serum creatinine monitoring during allopurinol therapy in ambulatory patients. Ann Pharmacother. 2006;40:386–91. doi: 10.1345/aph.1G589. [DOI] [PubMed] [Google Scholar]
- 24.Avanzini F, Tognoni G, Alli C, Colombo F, Herxheimer A. How informed general practitioners manage mild hypertension: a survey of readers of drug bulletins in 7 countries. International Society of Drug Bulletins (ISDB) Eur J Clin Pharmacol. 1996;49:445–50. doi: 10.1007/BF00195929. [DOI] [PubMed] [Google Scholar]
- 25.Sever P, Beevers G, Bulpitt C, Lever A, Ramsay L, Reid J, Swales J. Management guidelines in essential hypertension: report of the second working party of the British Hypertension Society. BMJ. 1993;306:983–7. doi: 10.1136/bmj.306.6883.983. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Campbell SM, Roland MO, Middleton E, Reeves D. Improvements in quality of clinical care in English general practice 1998–2003: longitudinal observational study. BMJ. 2005;331:1121. doi: 10.1136/bmj.38632.611123.AE. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Vrijens B, Vincze G, Kristanto P, Urquhart J, Burnier M. Adherence to prescribed antihypertensive drug treatments: longitudinal study of electronically compiled dosing histories. BMJ. 2008;336:1114–7. doi: 10.1136/bmj.39553.670231.25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Ashworth M, Medina J, Morgan M. Effect of social deprivation on blood pressure monitoring and control in England: a survey of data from the quality and outcomes framework. BMJ. 2008;337:a2030. doi: 10.1136/bmj.a2030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.The Oxford English Dictionary. Online). Available at http://dictionary.oed.com (last accessed 17 July 2009.