The use of cardiovascular risk factor information in practice databases: making the best of patient data

Tom Marshall

Abstract

Background Primary care teams record cardiovascular risk factor data on their patients to help them identify and treat patients eligible for prevention. However, it is not known to what extent this information is already available to clinicians, or the extent to which it is used.

Aim To assess the extent to which risk factor is recorded, and to determine the cost-effectiveness of using recorded risk factor information in order to identify and treat eligible patients.

Design of study An Excel-based model of the incremental costs and benefits of assessment and treatment.

Setting Two general practices in the West Midlands.

Method Untreated, non-diabetic patients, aged 35–74 years, were identified from each practice, and risk factor data was uploaded into an Excel spreadsheet. The completeness of risk factor data was assessed. The costs and benefits of assessing and treating patients, in descending order of estimated cardiovascular risk, were then modelled.

Results In each practice, 72.9% and 77.7% of patients had a record of their blood pressure, 26.9% and 25.7% were eligible for at least one treatment: aspirin was the most common treatment followed by antihypertensives. With patients systematically assessed in descending order of cardiovascular risk, 78% of eligible patients and 87% of preventable cardiovascular events are found in the first two deciles of the target population.

Conclusions Lack of risk factor information is not the principal constraint on cardiovascular prevention. Practices have sufficient risk factor data to inform an efficient, targeted prevention strategy.

INTRODUCTION

Cardiovascular disease (CVD) prevention requires identification and treatment of patients at high risk. Estimating CVD risk and eligibility for treatment necessitates clinicians to determine each patient's risk factor status for age, sex, diabetic status, smoking status, blood pressure and lipid levels. UK guidelines recommend antihypertensive treatment for those whose blood pressures exceed 160/100 mmHg, or with blood pressures exceeding 140/90 mmHg and 10-year CVD risk over 20%; aspirin is recommended for those at over 20% 10-year CVD risk who are aged over 50 years.1 The most recently available UK guidelines for statins recommend their use in all patients at over 30% 10-year coronary risk, or with familial hyperlipidaemia.2

In the UK, assessing patients' risk factor status has been the focus of government policy for some years. This is more usually undertaken opportunistically than systematically. However, in many cases risk factor data are already recorded in practice databases. Every patient's age and sex is known. Diabetic status is usually known — either from a practice-based diabetic register, or inferred from prescribing records. Increasingly, practices also have a record of patients' blood pressure, smoking status and cholesterol levels. Given that much information is already known, the question arises: are these data being used to their full potential? In particular are there sufficient data to inform a systematic approach to CVD prevention?

This paper reports an analysis of CVD risk factor data recorded on untreated patients, aged 35–74 years, in two West Midlands' practices. It determines how many patients are probably eligible for treatment, on the basis of recorded risk factor information, and how much CVD is preventable in these patients. It then estimates the costs and benefits of identifying and treating patients using a systematic compared to an opportunistic approach.

METHOD

Data sources

As part of a pilot project to help identify untreated patients eligible for CVD prevention, data were obtained from two practices (one urban and one semi-rural) in the West Midlands. In late 2002 (Practice 2) and early 2003 (Practice 1) an electronic search of the practice database was conducted to identify the target population. This included all patients aged 35–74 years, without cardiovascular disease or diabetes, and excluded patients currently on antihypertensive drug treatment or statins.

For each patient the following risk factors were extracted from the practice database: age, sex, smoking status, any recorded blood pressures in the past year, any recorded cholesterol levels in the past year, and the presence of left ventricular hypertrophy on electrocardiogram (ECG).

Data were validated by comparing the average recorded blood pressure, total to HDL cholesterol ratio and prevalence of smoking in each practice against the equivalent population drawn the Health Survey for England 1998.3

Determining risk and probable eligibility for treatment

Where a risk factor was not known, a default (or prior estimate) of the risk factor status was used instead. It is assumed that a patient is: a non-smoker, if smoking status is unknown; without left ventricular hypertrophy, if no ECG results are available; and, has blood pressure and cholesterol levels that are average for a person of their age and sex, if these are also unknown. This follows a previously described methodology.4 Because a nationally representative survey provides a larger and more representative sample than measurements taken within any individual practice, the default blood pressure and cholesterol levels were derived from the Health Survey for England of 1998.3

Risk factor data were entered into an Excel spreadsheet. Ten-year coronary heart disease (CHD) and cerebrovascular accident (CVA) risk were calculated for each patient.5^–7 Treatment eligibility was then determined for each patient using logical functions. Where more than one blood pressure or cholesterol measurement was available, the average of all the available measurements was used to calculate risk and determine treatment eligibility.

From this the proportions of patients with complete information on each risk factor, and the numbers of persons who can be identified as probably eligible for treatment, on the basis of recorded risk factor information, were calculated. Eligibility criteria are those described above. Any patient with total to HDL cholesterol ratio 8.0 or above is considered to have familial hypercholesterolaemia.

How this fits in

Primary care teams are encouraged to collect risk factor data for the purposes of identifying and treating eligible patients. Practice databases already contain sufficient cardiovascular risk factor information to identify patients who can benefit from preventive interventions. Typically, one quarter of patients, aged 35–74 years and not currently on treatment, are eligible for aspirin, antihypertensives or statins. With an appropriate systematic strategy, the vast majority of eligible patients could be identified by assessing only 20% of patients aged 35–74 years not currently on treatment. Primary care teams could make better use of information technology and recorded data to target cardiovascular disease prevention.

Estimating benefits of treatment

Both costs and benefits are estimated on the assumption that all patients accept and adhere to treatment. Benefits of treatment were estimated by multiplying 10-year CHD risk and 10-year CVA risk by the relative risk on treatment. All benefits were discounted at three per cent per year.

Estimates of effectiveness of aspirin and statins were derived from systematic reviews.8^–10 Risk reduction with statins is one-third of the full effect in the first year of use, two-thirds in the second and fully effective in the third.11

The 10-year absolute risk reduction with aspirin is reduced by 0.6% to take account of the increased risk of major bleeds.8 An estimate of the effects of each antihypertensive drug was derived as follows. The effect of treatment on systolic blood pressure was calculated using the results of meta-analysis.12 For example, the effect of bendroflumethazide, 2.5 mg on pre-treatment systolic blood pressure of 154 mmHg, is an 8.8 mmHg reduction. The effect is 1.0 mmHg greater on pre-treatment blood pressure that is 10 mmHg higher. The relative risks of CHD and CVA are 0.83 and 0.64, respectively, for each 12 mmHg reduction in blood pressure.13 This is consistent with the epidemiological evidence.14^,15 The relative risk of CHD following a 24 mmHg reduction in systolic blood pressure would therefore be 0.83^(24/12) (0.83² = 0.67). The effects of two, three and four antihypertensive drugs can be calculated from their effect on systolic blood pressure (see Table 1).

View this table:

Table 1

Effectiveness of treatments to prevent cardiovascular disease.

The number of CVD events prevented is the sum of the individual reductions in absolute risk of CVD.

Estimating costs of treatment

Costs of treatment were estimated on the following basis. It is assumed that the process of ranking patients takes 2 hours of GP time. Individual patient assessment prior to treatment is assumed to require an hour of practice nurse time and two cholesterol estimations. Annual follow up of patients on treatment requires half an hour (two clinic visits) of practice nurse time and half an hour of GP time; annual electrolytes and lipid levels; and annual liver function tests if taking statins. Costs of staff time were obtained from published sources: £32 and £118 per hour for practice nurse and GP clinic time, respectively.16

Drugs used were aspirin 75 mg, simvastatin 40 mg and antihypertensive drugs. Patients are assumed to require two, three or four antihypertensive drugs depending on their pre-treatment systolic blood pressure. Blood pressures under 160 mmHg systolic require bendroflumethazide 2.5 mg and metoprolol 100 mg; blood pressures 160 to 170 mmHg also require enalapril 10 mg; blood pressures over 170 mmHg also require amlodipine 5 mg. Drug costs were obtained from the British National Formulary.17 Dispensing costs are calculated at £0.85 per prescribed item on the assumption that four prescriptions are issued per year for long-term medication.18 All costs are discounted at 3% annually.

Efficiency of systematic versus opportunistic strategies

To model a systematic strategy informed by risk factor data in the practice database, all patients in the population are ranked by their estimated 10-year CVD risk (highest first). This represents the order in which patients would be systematically invited for cardiovascular risk assessment in a strategy informed by risk factor data. This list is then divided into deciles. The number of persons eligible for treatment in each decile is determined. The number of cardiovascular events that would be prevented by assessing, treating and following up these patients is determined: this is the sum of the individual reductions in 10-year CVD risk. The costs of prevention, for each decile, are the sum of the costs of assessment, treatment and follow up. Cost-effectiveness is calculated as cost per CVD event prevented, for each decile. In an opportunistic strategy, patients in the target population are effectively seen in random order. The number of persons eligible for treatment, costs, benefits and cost-effectiveness are, therefore, the same for each decile.

RESULTS

In Practice 1, 2866 (47%) patients, from a total list of 6100, were aged 35–74 years without diabetes and eligible for primary prevention. In Practice 2, there were 3877 (52%), from a list size of 7400.

Numbers of patients with complete risk factor information

In both practices, there was a record of smoking status for every patient. In Practice 1, 72.9% of patients had at least one record of blood pressure, and 21.6% at least one cholesterol level. In Practice 2, these figures were 77.7 and 16.5%, respectively. In Practice 2 (which provided all blood pressures for the previous 3 years) 44.2% of patients had at least two records of blood pressures (see Table 2).

View this table:

Table 2

Availability of risk factor information in two practices.

Validation of data

In the two practices overall average blood pressure was 135/83 mm Hg; average total to HDL cholesterol ratio was 4.3; smoking had a prevalence of 20.7%. In 5481 equivalent patients aged 35–74 years from the Health Survey for England average blood pressure was 135/77 mm Hg; average total to HDL cholesterol ratio was 4.3; smoking prevalence was 22.6%.

Numbers of patients probably eligible for treatment

In Practice 1 26.9% and in Practice 2 25.7% of patients, in the target population, were probably eligible for at least one treatment. In both practices aspirin was the treatment for which patients were most commonly eligible, followed by antihypertensive treatment (see Table 2). Untreated patients, with recorded systolic blood pressures over 160 mm Hg, accounted for 6.8% (263), of the target population, inPractice 1 and 6.7% (192) in Practice 2. In both practices there were untreated patients with systolic blood pressures greater than 180 mm Hg: 0.8% (22), in Practice 1, and 1.7% (38), in Practice 2.

Efficiency of systematic versus opportunistic strategies

In Practice 1, if patients are seen in random order, an average of 104 from every decile (387 patients) will be eligible for treatment. Assessment, treatment and follow up will prevent 11.1 cardiovascular events, and will cost £158 283 per 10 years for each decile: £14 300 per CVD event prevented. Table 3 shows the effect of systematic assessment of patients in descending order of CVD risk. Under this strategy, all of the first two deciles assessed are eligible for at least one treatment. By assessing the first two deciles, the strategy identifies 74% of the patients eligible for at least one treatment, and 86% of the preventable CVD events. After the fourth decile has been assessed, virtually no further CVD can be prevented by further assessment and treatment.

View this table:

Table 3

Numbers of patients eligible for treatment, number of cardiovascular events prevented, costs and cost-effectiveness of a systematic strategy in Practice 1.

In Practice 2, if patients are seen in random order, 74 from every decile (286 patients) will be eligible for treatment. Assessment, treatment and follow up will prevent 6.5 cardiovascular events and will cost £108 197 per 10 years for each decile: £16 700 per CVD event prevented. Under a systematic strategy, all of the first two deciles assessed are eligible for at least one treatment. By assessing the first two deciles, the strategy identifies 78% of the patients eligible for at least one treatment, and 87% of the preventable CVD events. Virtually no further CVD can be prevented after the fourth decile has been assessed and treated (see Table 4).

View this table:

Table 4

Numbers of patients eligible for treatment, number of cardiovascular events prevented, costs and cost-effectiveness of a systematic strategy in Practice 2.

Incremental analysis

The method of analysis allows the incremental costs and incremental benefits of each intervention to be calculated. It also permits changes to be made to assumptions relating to follow up, treatment eligibility, uptake of treatment and effectiveness. In Practice 1 and Practice 2 follow-up costs are 57 and 56%, respectively, of the cost of assessing and treating the two highest risk deciles under the baseline assumptions. This cost can be halved by using a practice nurse for follow up, or by reducing follow-up frequency. In Practice 1, the fourth antihypertensive (amlodipine) accounts for only 2% of the benefits, but 8% of the costs. In Practice 2, the figures are 2 and 11%, respectively. In Practice 1, simvastatin accounts for 6% of benefits, but 17% of costs; in Practice 2, these figures are 5 and 10%, respectively.

DISCUSSION

Summary of main findings

Both practices have electronic records of the great majority of their patients' blood pressures. Cholesterol levels are recorded in only a minority of patients. Recorded data are sufficient to identify that onequarter of untreated patients in the target population are eligible for at least one preventive treatment. In both practices, existing recorded data are sufficient to inform a systematic CVD prevention strategy. By assessing only 20% of the target population, such a strategy could identify three quarters of patients eligible for preventive treatments and prevent over 85% of avoidable CVD events. Identifying patients at random (opportunistically) it would be necessary to assess 85% of the target population to achieve the same results.

Strengths of the study

Recorded risk factor data was obtained from the clinical databases of two practices and is, therefore, routinely available to the practices concerned. Levels of risk factor data recorded, and numbers of patients eligible for treatment, were very similar in both practices. Practices certainly vary in the extent to which they record risk factor information, but there is no reason to suppose these practices are atypical.

Limitations of the study

The study was undertaken prior to the new General Medical Services contract in April 2004. This provides specific incentives for recording of blood pressures in patients aged 45 years and over.19 Recording of risk factor information is, therefore, likely to have improved since the date of this study.

The study is only able to determine probable eligibility for treatment, because eligibility for treatment is estimated from incomplete data of fewer than the required number of blood pressure measurements. Some patients without a recorded blood pressure or cholesterol measurement have unusually high or unusually low blood pressures or cholesterol levels. Those with unusually high levels are eligible for treatment on this basis alone. Using an average blood pressure or cholesterol level, may therefore underestimate the numbers eligible for antihypertensives, or statins. However, because other risk factors are known, using an average blood pressure or cholesterol level makes less difference to the estimation of cardiovascular risk. This means that the ranking of patients by cardiovascular risk is probably a reasonably accurate way in which to estimate benefits, eligibility for aspirin and (in most cases) statins. Patients in the highest risk deciles are, therefore, likely to be largely correctly identified, but a greater proportion of them may be eligible for antihypertensive treatment.

Risk factor information may not be completely accurate. The prevalence of the number of smokers in Practice 2 (27%) is comparable to that in the Health Survey for England in the same age groups (26%); however, the prevalence in Practice 1 (16%) is considerably lower.3 Nevertheless, these are the data on which clinical decisions will be made.

Aspirin and, since 2004, simvastatin are available over the counter. Although, likely to be small in number, some patients taking these drugs may not have this information recorded in the practice database. The true costs and benefits are influenced by uptake of, and adherence to, treatment and any follow-up costs. These may vary from the analysis above, however, unless those at high risk are systematically less likely to accept or adhere to medication, a strategy of systematically targeting patients at high cardiovascular risk remains more efficient.

Costs

Using the assumptions described above about half of the total cost is attributable to follow up (largely staff costs) and just over a tenth is attributable to assessment. If follow up is carried three times a year by a practice nurse, follow-up cost is much lower and assessment costs are a larger proportion of the total. Similarly, if patient adherence rates decline, the costs of treatment and follow up are reduced and assessment costs are a larger proportion of the total. In either case this increases the efficiency of a systematic compared to an untargeted assessment and prevention strategy, strengthening the case for systematic prevention. It is possible that a systematic strategy includes additional costs that were not accounted for in this analysis, but even additional costs of £5 000 per practice do not affect the efficiency of targeted prevention.

Implications for future research

Not all patients who are eligible are treated. Possible explanations include: failure to return for assessment following a high blood pressure measurement; being offered but declining treatment; contraindications to treatment; physician recommendations differing from the guidelines. It is clear, however, that lack of risk factor information is not the principal obstacle to the uptake of preventive treatment. This casts doubt on a policy that encourages collection of more risk factor information. Policies that facilitate systematic invitation and assessment of patients at high risk may be more productive.

At present the process of extracting data, calculating cardiovascular risk and ranking patients is time consuming, however it should be a relatively simple for makers of database software to automate this task.

The results of this modelling study should be investigated empirically. It is theoretically possible to identify and target individuals for cardiovascular risk prevention. It should be of great interest to those involved in primary care to investigate the extent to which: patients attend for assessment if invited; predicted eligibility for treatment is confirmed; and treatments are accepted and adhered to.

Notes

Competing interests

The author has stated that there are none.

Received November 25, 2004.
Revision received March 16, 2005.
Accepted February 9, 2006.

REFERENCES

↵
1. Williams B,
2. Poulter NR,
3. Brown MJ,
4. et al.
(2004) Guidelines for management of hypertension: report of the fourth working party of the British Hypertension Society 2004-BHS IV. J Hum Hypertens 18:139–185.
OpenUrl CrossRef PubMed
↵
1. Wood D,
2. Durrington P,
3. Poulter N,
4. et al.
(1998) on behalf of the Societies. Joint British recommendations on prevention of coronary heart disease in clinical practice. Heart 80(Suppl 2):S1–S29.
OpenUrl FREE Full Text
↵
1. Department of Health
(2000) Health Survey for England 1998 (The Stationery Office, London).
↵
1. Marshall T,
2. Rouse A
(2002) Resource implications and health benefits of primary prevention strategies for cardiovascular disease in people aged 30 to 74: mathematical modelling study. BMJ 325:197–199.
OpenUrl Abstract/FREE Full Text
↵
1. Anderson KM,
2. Odell PM,
3. Silson PWF,
4. Kannel WB
(1991) Cardiovascular disease risk profiles. Am Heart J 121:293–298.
OpenUrl CrossRef PubMed
1. Anderson KA,
2. Wilson PWF,
3. Odell PM,
4. Kannel WB
(1991) An updated coronary risk profile. A statement for health professionals. Circulation 83:356–362.
OpenUrl FREE Full Text
↵
1. Wolf PA,
2. D'Agostino RB,
3. Belanger AJ,
4. Kannel WB
(1991) Probability of stroke: a risk profile from the Framingham study. Stroke 22:312–318.
OpenUrl Abstract/FREE Full Text
↵
1. Hayden M,
2. Pignone M,
3. Phillips C,
4. Mulrow C
(2002) Aspirin for the primary prevention of cardiovascular events: a summary of the evidence for the U.S. Preventive Services Task Force. Ann Intern Med 136(2):161–172.
OpenUrl CrossRef PubMed
1. LaRosa JC,
2. He J,
3. Vupputuri S
(1999) Effect of statins on risk of coronary disease: a meta-analysis of randomized controlled trials. JAMA 282(24):2340–2346.
OpenUrl CrossRef PubMed
↵
1. Warshafsky S,
2. Packard D,
3. Marks SJ,
4. et al.
(1999) Efficacy of 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors for prevention of stroke. J Gen Intern Med 14(12):763–774.
OpenUrl CrossRef PubMed
↵
1. Law MR,
2. Wald NJ,
3. Rudnicka AR
(2003) Quantifying effect of statins on low density lipoprotein cholesterol, ischaemic heart disease, and stroke: systematic review and meta-analysis. BMJ 326:1423–1426.
OpenUrl Abstract/FREE Full Text
↵
1. Law MR,
2. Wald NJ,
3. Morris JK,
4. Jordan RE
(2003) Value of low dose combination treatment with blood pressure lowering drugs: analysis of 354 randomised trials. BMJ 326:1427–1432.
OpenUrl Abstract/FREE Full Text
↵
1. Gueyffier F,
2. Boutitie F,
3. Boissel JP,
4. et al.
(1997) Effect of antihypertensive drug treatment on cardiovascular outcomes in women and men. A meta-analysis of individual patient data from randomized, controlled trials. The INDANA Investigators. Ann Intern Med 126(10):761–767.
OpenUrl CrossRef PubMed
↵
1. MacMahon S,
2. Peto R,
3. Cutler J,
4. et al.
(1990) Blood pressure, stroke, and coronary heart disease. Part 1, prolonged differences in blood pressure: prospective observational studies corrected for the regression dilution bias. Lancet 335(8692):765–774.
OpenUrl CrossRef PubMed
↵
1. Collins R,
2. Peto R,
3. MacMahon S,
4. et al.
(1990) Blood pressure, stroke, and coronary heart disease. Part 2, short-term reductions in blood pressure: overview of randomised drug trials in their epidemiological context. Lancet 335:827–839.
OpenUrl CrossRef PubMed
↵
1. Netten A,
2. Curtil L
(2000) Unit costs of health and social care 2000 (Personal Social Services Research Unit, University of Kent at Canterbury).
↵
1. British Medical Association, Royal Pharmaceutical Society of Great Britain
(2004) British National Formulary (BMA, London), 48.
↵
1. Prescription pricing authority
(2004) NHS Electronic Drug Tariff. http://www.ppa.org.uk/edt/November_2004/mindex.htm (accessed 29 Jun 2006).
↵
1. The Department of Health
(2003) Investing in general practice: the new general medical services contract. Appendix A (DoH, London) http://www.dh.gov.uk/assetRoot/04/08/27/10/04082710.pdf (accessed 29 Jun 2006).

In this issue

Download PDF

Email Article

Citation Tools

Keywords

More in this TOC Section

Show more Original Papers

Cited By...

Intended for Healthcare Professionals

Tweets by @BJGPjournal

[1] ↵
Williams B,
Poulter NR,
Brown MJ,
et al.
(2004) Guidelines for management of hypertension: report of the fourth working party of the British Hypertension Society 2004-BHS IV. J Hum Hypertens 18:139–185.
OpenUrl CrossRef PubMed

[2] Williams B,

[3] Poulter NR,

[4] Brown MJ,

[5] et al.

[6] ↵
Wood D,
Durrington P,
Poulter N,
et al.
(1998) on behalf of the Societies. Joint British recommendations on prevention of coronary heart disease in clinical practice. Heart 80(Suppl 2):S1–S29.
OpenUrl FREE Full Text

[7] Wood D,

[8] Durrington P,

[9] Poulter N,

[10] et al.

[11] ↵
Department of Health
(2000) Health Survey for England 1998 (The Stationery Office, London).

[12] Department of Health

[13] ↵
Marshall T,
Rouse A
(2002) Resource implications and health benefits of primary prevention strategies for cardiovascular disease in people aged 30 to 74: mathematical modelling study. BMJ 325:197–199.
OpenUrl Abstract/FREE Full Text

[14] Marshall T,

[15] Rouse A

[16] ↵
Anderson KM,
Odell PM,
Silson PWF,
Kannel WB
(1991) Cardiovascular disease risk profiles. Am Heart J 121:293–298.
OpenUrl CrossRef PubMed

[17] Anderson KM,

[18] Odell PM,

[19] Silson PWF,

[20] Kannel WB

[21] Anderson KA,
Wilson PWF,
Odell PM,
Kannel WB
(1991) An updated coronary risk profile. A statement for health professionals. Circulation 83:356–362.
OpenUrl FREE Full Text

[22] Anderson KA,

[23] Wilson PWF,

[24] Odell PM,

[25] Kannel WB

[26] ↵
Wolf PA,
D'Agostino RB,
Belanger AJ,
Kannel WB
(1991) Probability of stroke: a risk profile from the Framingham study. Stroke 22:312–318.
OpenUrl Abstract/FREE Full Text

[27] Wolf PA,

[28] D'Agostino RB,

[29] Belanger AJ,

[30] Kannel WB

[31] ↵
Hayden M,
Pignone M,
Phillips C,
Mulrow C
(2002) Aspirin for the primary prevention of cardiovascular events: a summary of the evidence for the U.S. Preventive Services Task Force. Ann Intern Med 136(2):161–172.
OpenUrl CrossRef PubMed

[32] Hayden M,

[33] Pignone M,

[34] Phillips C,

[35] Mulrow C

[36] LaRosa JC,
He J,
Vupputuri S
(1999) Effect of statins on risk of coronary disease: a meta-analysis of randomized controlled trials. JAMA 282(24):2340–2346.
OpenUrl CrossRef PubMed

[37] LaRosa JC,

[38] He J,

[39] Vupputuri S

[40] ↵
Warshafsky S,
Packard D,
Marks SJ,
et al.
(1999) Efficacy of 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors for prevention of stroke. J Gen Intern Med 14(12):763–774.
OpenUrl CrossRef PubMed

[41] Warshafsky S,

[42] Packard D,

[43] Marks SJ,

[44] et al.

[45] ↵
Law MR,
Wald NJ,
Rudnicka AR
(2003) Quantifying effect of statins on low density lipoprotein cholesterol, ischaemic heart disease, and stroke: systematic review and meta-analysis. BMJ 326:1423–1426.
OpenUrl Abstract/FREE Full Text

[46] Law MR,

[47] Wald NJ,

[48] Rudnicka AR

[49] ↵
Law MR,
Wald NJ,
Morris JK,
Jordan RE
(2003) Value of low dose combination treatment with blood pressure lowering drugs: analysis of 354 randomised trials. BMJ 326:1427–1432.
OpenUrl Abstract/FREE Full Text

[50] Law MR,

[51] Wald NJ,

[52] Morris JK,

[53] Jordan RE

[54] ↵
Gueyffier F,
Boutitie F,
Boissel JP,
et al.
(1997) Effect of antihypertensive drug treatment on cardiovascular outcomes in women and men. A meta-analysis of individual patient data from randomized, controlled trials. The INDANA Investigators. Ann Intern Med 126(10):761–767.
OpenUrl CrossRef PubMed

[55] Gueyffier F,

[56] Boutitie F,

[57] Boissel JP,

[58] et al.

[59] ↵
MacMahon S,
Peto R,
Cutler J,
et al.
(1990) Blood pressure, stroke, and coronary heart disease. Part 1, prolonged differences in blood pressure: prospective observational studies corrected for the regression dilution bias. Lancet 335(8692):765–774.
OpenUrl CrossRef PubMed

[60] MacMahon S,

[61] Peto R,

[62] Cutler J,

[63] et al.

[64] ↵
Collins R,
Peto R,
MacMahon S,
et al.
(1990) Blood pressure, stroke, and coronary heart disease. Part 2, short-term reductions in blood pressure: overview of randomised drug trials in their epidemiological context. Lancet 335:827–839.
OpenUrl CrossRef PubMed

[65] Collins R,

[66] Peto R,

[67] MacMahon S,

[68] et al.

[69] ↵
Netten A,
Curtil L
(2000) Unit costs of health and social care 2000 (Personal Social Services Research Unit, University of Kent at Canterbury).

[70] Netten A,

[71] Curtil L

[72] ↵
British Medical Association, Royal Pharmaceutical Society of Great Britain
(2004) British National Formulary (BMA, London), 48.

[73] British Medical Association, Royal Pharmaceutical Society of Great Britain

[74] ↵
Prescription pricing authority
(2004) NHS Electronic Drug Tariff. http://www.ppa.org.uk/edt/November_2004/mindex.htm (accessed 29 Jun 2006).

[75] Prescription pricing authority

[76] ↵
The Department of Health
(2003) Investing in general practice: the new general medical services contract. Appendix A (DoH, London) http://www.dh.gov.uk/assetRoot/04/08/27/10/04082710.pdf (accessed 29 Jun 2006).

[77] The Department of Health

Main menu

User menu

Search

The use of cardiovascular risk factor information in practice databases: making the best of patient data

Abstract

INTRODUCTION

METHOD

Data sources

Determining risk and probable eligibility for treatment

How this fits in

Estimating benefits of treatment

Estimating costs of treatment

Efficiency of systematic versus opportunistic strategies

RESULTS

Numbers of patients with complete risk factor information

Validation of data

Numbers of patients probably eligible for treatment

Efficiency of systematic versus opportunistic strategies

Incremental analysis

DISCUSSION

Summary of main findings

Strengths of the study

Limitations of the study

Costs

Implications for future research

Notes

Competing interests

REFERENCES

In this issue

Citation Manager Formats

Jump to section

Keywords

More in this TOC Section

Related Articles

Cited By...

NAVIGATE

RCGP

MY ACCOUNT

NEWS AND UPDATES

AUTHORS & REVIEWERS

CUSTOMER SERVICES

CONTRIBUTE

CONTACT US