Introduction
Hypertension (HT) is a recognized cardiovascular risk factor (CRF) which absorbs a large part of available primary care resources1-3. For most patients with HT, the aim of efforts to control blood pressure (BP) is to maintain figures below 140/90 mmHg. In patients with HT who also have diabetes, the target figures should be even lower, as recent clinical trials have shown4,5. Accordingly, the figures for these patients have been set at 130/85 mmHg6,7, although some authorities recommend even lower values8,9. The difference is due to the greater cardiovascular risk in patients who also have diabetes. Because many micro- and macroangiopathic complications that occur in patients with diabetes are aggravated by HT, blood pressure in these persons needs to be controlled much more strictly4.
It is currently accepted that because of the chronic nature and high prevalence of HT, health care aimed at controlling this disorder should be provided mainly by primary care centers (PCC)1,2. In our setting, however, the situation varies widely: some PCCs are located in rural areas, others are in urban areas; some are teaching centers, others are not. Might these differences lead to variations in the results of continued care for these patients?
The aim of this study was to determine whether BP was adequately controlled in the population with HT alone and with diabetes served by PCCs in Catalonia. We also investigated whether the characteristics of the center influenced the degree of control of HT in these populations.
Material and methods
Design
This descriptive, cross-sectional, multicenter study involved 31 PCC in Catalonia and was done in 1998. The clinical data we studied were from the period from January to December 1996, one year after the publication of practical guidelines for the management of HT in Catalonian PCCs10.
Study population
The region of Catalonia is divided into four primary care subdivisions administered by the Catalonian Institute of Health. Eighteen percent of the centers are located in the Barcelona Nord-Maresma-Girona subdivision; 32% in Centre-Lleida, 38% in Costa de Ponent-Tarragona-Tortosa, and 12% in Barcelona. Seventy percent of the centers are not involved in teaching activities, and the other 30% are teaching centers. The centers were sampled randomly according to a stratified scheme that reproduced the relative percentages of centers in each subdivision. Centers where implementation of the most recent reforms was pending were excluded from this study. Table 1 summarizes the numbers and distribution of the centers that participated.
In the second stage, clinical records held at the centers were sampled randomly, taking into account the total number of patients with HT seen at each center. The sample size necessary to fulfil the requirements of a 95% confidence interval and a 3% degree of precision was calculated, and mean expected prevalence of HT in the population served by all centers was estimated at 12%.
Data collection
Six persons were trained in a pilot study to audit the clinical records, which were chosen for study just before the data were recorded. A pilot test with the same procedure as was later used for the centers actually included in the study was done with 50 charts from 10 centers considered representative of the region of Catalonia. Problems identified in the pilot study led to some modifications in the questionnaire to adapt it to the information contained in the chart.
These 6 persons audited the clinical records used in all PCCs. None of the auditors had any personal or professional relationship with any of the participating centers.
The following variables were recorded:
Characteristics of the center.
Number of years in operation, teaching or non-teaching, rural or urban. A rural center was defined as a local center separate from the PCC, required to serve a widely dispersed population. A teaching center was defined as one where resident physicians worked in addition to permanent staff.
Physician characteristics.
Age, sex, specialty, type of contract, number of patients, and hours spent per day seeing patients.
Patient characteristics.
Age, sex, known duration of HT, last two BP values, screening for and diagnosis of any of the following CRF: smoking (recorded in the medical record as smoker vs. nonsmoker), diabetes (two glucemia values > 140 mg/dL), obesity (body mass index >= 30) and hypercholesterolemia (two cholesterol values > 200 mg/dL). Screening was recorded when the record noted any screening procedure for CRF via interview, laboratory analysis or electrocardiogram. The degree of staff compliance in recording the indicators proposed by the Working Group on Hypertension of the Catalonian Society for Family and Community Medicine (evaluation of the diagnosis, follow-up, check-ups, multifactorial approach and structure was also evaluated10. The data from this part of the audit will be analyzed in a separate study.
Statistical analysis
The statistical analysis was done with the SPSS statistical package for Windows; 95% confidence intervals were used. The possible relationships between variables were analyzed with the chi-square test for qualitative variables, and with Student's t test for quantitative variables. Differences were considered significant at p values ¾ 0.05.
Results
Description of the sample
From all patients served by the 31 centers included in the study, we selected 2240 with hypertension, of whom 495 (22.1%) also had diabetes. Teaching centers were responsible for 34.6% of all patients with HT sampled for this study, and 42.1% were seen at rural centers. The center and patient characteristics are summarized in Tables 2 and 3.
Mean age of the physicians was 39.9 years, and 56.7% were men. Slightly more than three quarters (78.4%) were members of the permanent staff, and 43.7% were family physicians.
Mean age of the patients was 64.9 years (95% CI 64.4-65.4 years), and 61.2% were women. In the subgroup with both HT and diabetes, mean age was 67.7 years (95% CI 66.6-68.7 years), and 65% were women.
The data for CRF showed that 40.5% of the patients had at least one other risk factor in addition to HT; this was the largest of the subgroups of patients with one or more CRF. The most common CRF was hypercholesterolemia (54.5%), followed by obesity (47.6%).
Control of blood pressure
Mean systolic (SBP) and diastolic blood pressure (DBP) was 145.6 mmHg (95% CI 145-146 mmHg) and 84.8 mmHg (95% CI 84.4-85.2 mmHg) respectively. In the subgroup of patients with HT and diabetes, these figures were 147.9 mmHg (95% CI 147-149 mmHg) and 83.2 mmHg (95% CI 82.4-84.0 mmHg) respectively.
Optimum control of HT (BP < 140/90 mmHg) was attained in 25.7% of all cases (95% CI 23.9-27.5%). In the subgroup with HT and diabetes, BP was < 140/90 mmHg in 24.7% of the patients (95% CI 21.0-28.4%). The difference between patients with HT only and those with HT and diabetes was not significant. In the latter subgroup, the proportion of patients whose BP was < 130/85 mmHg decreased to 6.7% (95% CI 4.5-8.9%).
Variables associated with control of BP
Systolic BP was higher in patients older than 65 years and in patients followed at urban and teaching centers, although the difference was significant only for teaching centers (p = 0.0001) and patients older than 65 years (p = 0.0001). Diastolic BP was significantly higher in patients followed at urban centers (p = 0.0001)and teaching centers (p = 0.001). In contrast with systolic BP, which was higher in patients older than 65 years, diastolic BP was higher in patients younger than 65 years (p = 0.0001).
The relationships between BP values and different factors analyzed in this study are summarized in Table 4.
Blood pressure was higher in patients with more than one CRF in addition to diabetes, and in patients who were followed at centers that had been open for less than 6 years, although these differences were not statistically significant. We found no relation between physician's age and degree of control of HT.
The relationships between BP values and different factors we studied in the population of patients with HT and diabetes are summarized in Table 5. In this population, systolic BP was higher at teaching centers and at urban centers, in patients older than 65 years, and when one other CRF was present in addition to HT and diabetes. However, the difference was significant only for teaching centers (p = 0.005). Diastolic BP was higher in urban centers (statistically nonsignificant), in patients younger than 65 years (p = 0.001), and in patients followed at teaching centers (p = 0.036).
Discussion
Hypertension is a CRF whose control and follow-up are mainly the responsibility of the primary care physician1-3. Studies have shown that above the cut-off values used to diagnose HT and evaluate the effectiveness of control measures, the risk of a cardiovascular event is increased. These cut-off values are currently 140/90 mmHg in patients without diabetes, and 130/85 mmHg in patients with diabetes6,7.
Many studies11-14 have looked at the prevalence and degree of control of HT in Spain, although few of them have taken into account the influence of characteristics of the center responsible for the patient. The Controlpres-95 study11, which involved a total of 7032 patients with HT, investigated the degree of control of HT in Spain with a questionnaire that was used in primary care settings and outpatient cardiology clinics. In this study, the optimum control value (BP < 140/90 mmHg) was found in no more than 13% of all patients treated. When suboptimal control figures were considered (BP < 160/90 mmHg), this figure increased to 18%.
In the Controlpres-98 study12, which involved only primary care centers and recruited 6900 patients with HT, the proportion of patients with an optimum degree of control was somewhat higher: 16.3%. When PB values ¾ 140/90 mmHg were used, the percentage increased to 30.1%. In another study Banegas et al13 found that the percentage of treated patients in whom control of BP was considered adequate (< 140/90 mmHg) was only 15.5%. Plans et al14, in a study that was limited to patients with HT in Catalonia, found that BP was lower than 160/95 mmHg in 21% of the patients. This proportion is lower than that in the present study, especially in view of the higher cut-off value used by Plans et al. to consider BP control adequate. In the 7 years that separate these two studies, primary health care services have been reformed in Catalonia, team care has been fomented, and programs for the diagnosis and follow-up of HT have been established. These factors have undoubtedly helped to improve these percentages.
The proportion of patients with HT who were followed at PCCs in Catalonia and in whom BP could be considered optimally controlled (< 140/90 mmHg) was 25.7%, a figure somewhat higher than in earlier studies. In a study of a population in Hospitalet15, HT was considered well controlled in 36.4% of the patients. Although this figure is much higher than that in the present study, it should be noted that the BP value taken to indicate a good degree of control in the Hospitalet study was 160/90 mmHg.
We believe that our findings for the control of HT are better than those of other studies, as we investigated only health centers operating under the new system, where HT is managed according to a standardized protocol10,16. In the subgroup of patients with HT and diabetes, a population that received particular attention in this study, 24.7% of the patients had BP < 140/90 mmHg. As noted above, this cut-off does not reflect an adequate degree of control in this subgroup4,5, as current recommendation for these patients is to maintain BP below 130/85 mmHg. Only 6.7% percent of the patients with HT and diabetes in the present study met this stricter criterion, a proportion below the 10% recently reported by Hanninen et al17. However, these authors used a cut-off of 140/90 mmHg.
A further aim of our study was to compare the degree of control of BP between urban and rural centers, and between teaching and non-teaching centers. In teaching centers 10.8% of the patients were considered to have well controlled BP, whereas in non-teaching centers this proportion24.9%was significantly higher.
A search of the Medline database for items published between 1960 and 2000 failed to identify any studies that examined the influence of characteristics of the center where the patients were followed. None of the articles published during this period compared the influence of teaching vs. non-teaching centers on the degree of control of HT. Although studies have been published that investigated control of HT in rural populations18, none have compared rural vs. urban centers. In addition, we could not find any multicenter studies in Spain that mentioned the percentage of patients with HT and diabetes whose BP met currently recommended criteria. Surprisingly, at teaching centers, where activities are more often performed according to standard protocols to fulfil training requirements for residents, and where the degree of control of HT might be assumed to be higher, we found that the degree of control attained was worse than in non-teaching centers, both for all patients with HT and for the subgroup with HT and diabetes. In fact, the use of protocols to control cardiovascular risk is not accompanied
by improved control of CRF19, as a study in the Basque Country showed. Although the number of patients recorded as having CRF increases, control is not necessarily improved20. For similar reasons HT might be expected to be better controlled at urban centers because of the greater availability of resources; however, our findings show the opposite. We believe these differences may be caused by confounding factors which where not determined in the present study, such as differences in life style, eating habits or even the patient's degree of trust in his or her physician, and having higher BP at the time of diagnosis. Further studies will be needed to determine the causes of these differences.
The CRF most often associated with HT were obesity and hypercholesterolemia, a finding consistent with the results of the Controlpres-9511 and Controlpres-98 studies12. The percentage of patients with HT who smoked was only 12.4%, a figure similar to that found in these two earlier studies. In summary, we conclude that BP was decreased in a small proportion of patients with HT who were followed at PCC in Catalonia. The percentage was even smaller for patients with HT who also had diabetes (6.7%), a population in which HT should be controlled more effectively.
Spanish version available at www.atencionprimaria.com/42194 A commentary follows this article. (pág. 379) Equipo de Atención Primaria Gòtic. Institut Català de la Salut. Barcelona. Correspondence: Mencia Benítez Camps EAP Gótic. Passatje de la Pau, 1. 08002 Barcelona, Spain. Manuscript accepted for publication 13.06.01.
