Objective: To evaluate how well hypertension is managed inHMO patients and to assess opportunities for improvement.
Study Design: Retrospective cohort study.
Patients and Methods: The study population included HMOmembers (age 45-84 years) who had at least 1 ambulatoryencounter with an ICD-9-CM diagnosis code of essential hypertensionduring the first 6 months of 1999. Medical records werereviewed to obtain information on blood pressure measurements,sex, age, coexisting medical conditions, smoking status, andchanges made to the antihypertensive drug regimen.
Results: We identified 681 members with 3347 encountersrelated to hypertension management during 1999. Overall, 74(11%) patients were at target blood pressure for all visits and 260(38%) were at target blood pressure for at least 50% of the visits;222 (33%) patients were not at target blood pressure for any visit.A history of coronary artery disease or cerebrovascular disease wasassociated with better blood pressure control (defined as being atgoal levels during at least 50% of visits), while being older (age≥75) or having diabetes mellitus was associated with poorer control.Medication regimen intensifications occurred in 10% of visitswith systolic blood pressure levels of 140-149 mm Hg, comparedwith 45% of visits with levels of ≥180 mm Hg. Medication regimenintensifications occurred in 21% of visits with diastolic blood pressurelevels of 90-99 mm Hg and 43% of visits with levels of ≥100mm Hg.
Conclusion: Efforts are required to reduce "therapeutic inertia,"particularly in patients with modestly elevated systolic blood pressurelevels.
(Am J Manag Care. 2004;10:481-486)
Hypertension is among the most common outpatientdiagnoses in the United States.1 The accumulatedevidence establishing the benefits oftreating hypertension far exceeds that which exists forthe treatment of most medical conditions managed inthe ambulatory setting. Although primary care physiciansmay be aware of the existence of treatment guidelinesfor hypertension,2 these physicians' bloodpressure thresholds for the diagnosis and treatment ofhypertension reportedly are often substantially higherthan recommended.3 Data from the most recentNational Health and Nutrition Examination Survey(NHANES 1999-2000) indicate that hypertension prevalenceis increasing in the United States, and hypertensioncontrol rates are unacceptably low.4 In 1999-2000,28.7% of NHANES participants had hypertension, a 3.7%increase compared with 1988-1991.4
Control of high blood pressure has been included asa Health Plan Employer Data and Information Set(HEDIS®) measure by the National Committee forQuality Assurance as part of its program to compare theperformances of managed health care plans.5 Alexanderand colleagues previously demonstrated that it is feasibleto assess blood pressure control as a quality measurethrough review of medical records in an HMOpopulation.6 In the present study, we evaluated thequality of hypertension management in patients followedin a managed care setting in order to assess theopportunities that exist for improvement.
Study Population and Design
InternationalClassification of Diseases, Ninth Revision, ClinicalModification
A retrospective study was conducted among patientsenrolled in a mixed-model, not-for-profit HMO operatingin New England. The study population included a randomsample of 770 HMO members who were 45-84 yearsold as of January 1, 1999; had continuous plan enrollmentin 1999; and had at least 1 ambulatory encounterwith a diagnosis of essential hypertension ( [ICD-9-CM] code 401.xx) during the first 6months of 1999 documented in the HMO administrativedatabases. The age characteristics for study subjectsrelated to the HEDIS® 2000 measure "Controlling HighBlood Pressure." Medical records of these members werereviewed by trained nurse abstractors for all ambulatoryvisits to any healthcare provider that were related to themanagement of hypertension from January 1, 1998,through December 31, 1999. Information was obtainedon blood pressure measurements, demographic characteristics(sex and age), coexisting medical conditions(diabetes mellitus, coronary artery disease, cerebrovasculardisease, peripheral arterial disease, and hypercholesterolemia),smoking status, and changes made to theantihypertensive drug regimen because of the patient'sblood pressure level.
For the purpose of this study, diabetes mellitus wasdefined by use of a hypoglycemic agent, a documentedplasma glucose level of >200 mg/dL, or a glycosylatedhemoglobin level of >6%. Coronary artery disease wasconsidered to be present based on a diagnosis documentedin the medical record or a history of myocardialinfarction, angina, coronary artery bypass graft surgery,percutaneous transluminal coronary angioplasty, or prescriptionof nitrate therapy. Cerebrovascular disease wasdetermined based on a diagnosis in the medical record,or documentation of a history of stroke or transientischemic attack. Peripheral arterial disease was consideredpresent based on a diagnosis in the medical record.Hyperlipidemia was defined by use of a lipid-loweringagent or a total cholesterol level of ≥240 mg/dL.
The study population was further restricted toinclude members who had at least 1 ambulatory encounterassociated with the prescription of an antihypertensivemedication or a blood pressure measurementabove a specified target level documented in the medicalchart in both 1998 and 1999. Target blood pressurecontrol levels were defined as a systolic blood pressureof <140 mm Hg and a diastolic blood pressure of <90mm Hg for nondiabetic patients, and a systolic bloodpressure of <130 mm Hg and a diastolic blood pressureof <85 mm Hg for diabetic patients, consistent withguidelines in effect at the time of the study.7
When multiple blood pressure readings were documentedfor the same visit, the blood pressure readingwith the lowest systolic blood pressure level during thevisit was used in our analyses, regardless of the level ofthe diastolic blood pressure. For each visit, an intensificationof the antihypertensive regimen was considered tohave occurred if the dose of an antihypertensive medicationhad been increased, or a new agent was prescribed.
The numbers and proportions of patients with bloodpressure control for at least 50% of visits during calendaryear 1999 were determined–both overall and by patientdemographic characteristics (sex and age), coexistingmedical conditions (diabetes mellitus, coronary arterydisease, cerebrovascular disease, peripheral arterial disease,and hypercholesterolemia), smoking status, historyof blood pressure control (percentage of visits in the previousyear with blood pressure at target level), and historyof antihypertensive medication intensification(percentage of visits in the previous year with a medicationintensification). Based on the distribution of thedata, the percentages of visits in the previous year withblood pressure at target level were categorized as 0%, 1%-25%, 26%-50%, 51%-75%, and >75%; the percentages ofvisits in the previous year with medication intensificationwere categorized as 0%, 1%-20%, and >20%. Statistical significanceof differences was tested by using the Pearsonchi-square statistic and the Mantel-Haenszel test for linearassociation. Logistic regression was used to estimatethe strength of the association between selected patientcharacteristics and blood pressure at target level for atleast 50% of the visits. Stepwise logistic regression wasused to identify the multivariate predictors of bloodpressure at target level, with the criterion for entry intothe models being a probability value of <.05.
Using data from all visits for which the patient'sblood pressure was not at target level during calendaryear 1999, random effects logistic regression8 was usedto evaluate the associations between patient characteristicsand intensification of the antihypertensive regimen.Models were constructed that included variablesfor patient sex, age (<55 years, 55-64 years, 65-74years, and ≥75 years of age), coexisting medical conditions(diabetes mellitus, coronary artery disease, cerebrovasculardisease, peripheral arterial disease, andhypercholesterolemia), smoking status (current, pastor nonsmoker), history of blood pressure control (controlat previous visit and the percentage of visits in theprevious year with blood pressure at target level), historyof antihypertensive medication intensification(intensification at previous visit and the percentage ofvisits in the previous year with a medication intensification),the number of days since the last ambulatoryvisit related to management of high blood pressure, systolicblood pressure level (<130, 130-139, 140-149,150-159, and ≥160 mm Hg), and diastolic blood pressurelevel (<80, 80-84, 85-89, 90-99, and ≥100 mm Hg).
Medical records were located for 758 (98%) of the 770patients selected for study. Of these, 10 patients had nodocumentation of hypertension or use of antihypertensiveagents in the medical chart during 1998 and 1999,and an additional 67 patients did not have an ambulatoryvisit potentially related to hypertension in both calendaryears. Thus, 681 members, with 3347 encountersrelated to the management ofhypertension during the calendaryear 1999, met the criteriafor inclusion in the study. Themean number of visits per memberwas 4.9 (range 1 to 24) during1999, and 4.6 (range 1 to 24)during 1998. Among eligiblepatients, 637 (94%) had morethan 1 visit and 519 (76%) hadmore than 2 visits in 1999, and606 (89%) had more than 1 visitand 485 (71%) had more than 2visits in 1998. The mean age ofthe study population was 66years and 55% (n = 373) werefemale.
Overall, 74 (11%) patientswere at target blood pressure forall visits and 260 (38%) were attarget blood pressure for at least50% of visits during 1999; 222(33%) patients were not at targetblood pressure for any visit during1999. The relationshipbetween patient characteristicsand having been at target bloodpressure for at least 50% of visitsduring 1999 is shown in Table1. As detailed in Table 2, olderpatients were less likely to be attarget blood pressure (adjustedodds ratio [OR] = 0.49, 95% confidenceinterval [CI] = 0.29,0.84 among those age 75 yearsor older compared with thoseless than 55 years of age), aswere patients with a diagnosis ofdiabetes mellitus (adjusted OR= 0.32, 95% CI = 0.18, 0.56).Patients with a diagnosis ofcoronary artery disease andthose with a diagnosis of cerebrovasculardisease were morelikely to be at target blood pressurefor at least 50% of visits, aswere patients with a prior historyof blood pressure control.Patient sex, a diagnosis ofhypercholesterolemia, peripheralarterial disease, smoking status, and prior medicationchange history were not significantly associated withblood pressure control.
For 388 of the 3347 encounters (12%) during 1999,the patient had the antihypertensive therapeutic regimenintensified (dose increase or switch to a new agent).Medication intensification was more likely to occur inthe context of visits where the patient's blood pressurewas not under control. For 47 of the 1181 visits (4%)where the patient's blood pressure was at target level,the patient had a medication intensification. For 341 ofthe 2166 visits (16%) where the patient's blood pressurewas not at target level, the patient had a medicationincrease. Healthcare providers were more responsive tosubstantially elevated systolic and diastolic blood pressurelevels than they were to modestly elevated levels.Intensifications in the medication regimen occurred at10% of visits with systolic blood pressure levels of 140 to<150 mm Hg compared with 45% of visits with levels of180 mm Hg or more (Figure 1). Intensifications in themedication regimen occurred at 21% of visits with diastolicblood pressure levels of 90 to 99 mm Hg and 43% ofvisits with levels of 100 mm Hg or more (Figure 2).
Random effects logistic regression was used to examinethe associations between patient characteristics andantihypertensive therapy intensification, with data from2166 visits where the patient's blood pressure was notat target level (Table 3). The level of systolic blood pressureand the level of diastolic blood pressure were bothhighly predictive of antihypertensive therapy intensification.The unadjusted ORs associated with medicationintensification were 11.51 (95% CI = 6.08, 21.81) amongpatients who had a systolic blood pressure of ≥180 comparedwith those who had a systolic blood pressure of<140 mm Hg, and 24.82 (95% CI = 7.00, 88.05) amongpatients who had a diastolic blood pressure of ≥110 mmHg compared with those who had a diastolic blood pressureof <80 mm Hg. Men were more likely than womento have an intensification of antihypertensive therapy,as were those patients with a prior history of medicationchange. Patient age and coexisting medical conditionswere not significantly associated with a medicationincrease. There was a statistically significant interactionbetween gender and high diastolic blood pressure, andbetween gender and medication intensification at theprevious visit.
In full models estimated separately for men andwomen (Table 4), the levels of systolic blood pressureand diastolic blood pressure remained highly predictiveof medication intensification. For men, the adjusted ORsassociated with a medication intensification were 7.33(95% CI = 2.92,18.40) among patients who had a systolicblood pressure of ≥180 mm Hg compared with those whohad a systolic blood pressure of <140 mm Hg, and 15.10(95% CI = 2.25, 101.53) among patients who had a diastolicblood pressure of ≥110 mm Hg compared withthose who had a diastolic blood pressure of <80 mm Hg.For women, the adjusted ORs associated with a medicationintensification were 14.68 (95% CI = 5.03, 42.86)among patients who had a systolic blood pressure of≥180 mm Hg compared with those who had a systolicblood pressure of <140 mm Hg, and 7.27 (95% CI = 1.44,36.67) among patients who had a diastolic blood pressureof ≥110 mm Hg compared with those who had adiastolic blood pressure of <80 mm Hg. In both men andwomen, patients for whom more than 20% of visitsresulted in a medication intensification in the previousyear were more likely to have had a medication intensificationat the present visit. Blood pressure control atthe prior visit was only significantly related to presentvisit medication intensification in women.
The findings of our study serve to reinforce previousreports in the literature indicating that blood pressurecontrol rates are suboptimal.2,6,9,10 In the present study,deficiencies in the quality of hypertension managementwere observed despite the fact that patients wereassessed frequently, with an average of almost 5 visitsper year. According to a recently published report onover 400 indicators of healthcare quality, recommendedcare was not delivered to about a third of patientswith hypertension.11 While confirming deficiencies inthe quality of hypertension management in the UnitedStates, that report also brought attention to the verylarge number of areas that the primary care physicianmust address in providing care to patients in the ambulatorysetting. According to data from the NationalAmbulatory Medical Care Survey, the length of an averageoffice visit to a primary care physician is about 15minutes. Although more than 70% of office visits to primarycare physicians include blood pressure measurements,12 hypertension management competes againstan increasing array of preventive care priorities andother acute and chronic medical issues.
Oliveria and colleagues have examined barriers to primarycare physicians' willingness to increase the intensityof treatment among patients with uncontrolledhypertension.2 In that study, the most frequently citedreason for not addressing uncontrolled hypertensionrelated to satisfaction with the current blood pressurelevel, even if it was above the threshold level for treatment.On average, physicians reported that 150 mm Hgwas the lowest systolic blood pressure at which theywould recommend pharmacologic therapy to a patientwithout comorbidities. In the present study, intensificationsin the antihypertensive regimen were significantlymore likely to occur only when the systolic blood pressurelevel reached 160 mm Hg or greater. We also foundthat patients who had prior medication increases weremore likely to have an intensification in therapy. Priorefforts to modify the medication regimen may reduce anysubsequent reluctance to making additional changes.
One limitation of our study is the use of medicalrecords. Incomplete documentation of drug-prescribingdecisions may have led to an underestimation of medicationintensifications. In addition, incomplete documentationof diagnoses or laboratory test results may have ledto underascertainment of comorbidities. Information onthe reasons for the lack ofresponse by providers to elevatedblood pressure levels, as wellas provider characteristics, alsowas not available.
Our findings serve to emphasizethe opportunities that existto improve the care of patientswith hypertension in the ambulatorysetting. The development ofsuccessful strategies to improvethe quality of hypertension managementhinges on an enhancedunderstanding of the process ofhow physicians make decisionsabout initiating and modifyingpharmacologic treatment.13,14Special efforts are clearlyrequired to reduce therapeuticinertia,15 particularly in regardto modestly elevated systolicblood pressure levels, as thebenefits of blood pressurereduction can be substantial.16
From the Meyers Primary Care Institute, Fallon Foundation, and University ofMassachusetts Medical School, Worcester, Mass (JHG, SEA, TSF); Fallon Clinic, Inc,Worcester, Mass (MK, RB); Department of Medicine, University of Alberta, Edmonton,Alberta, Canada (SRM); and Department of Medicine, University of Massachusetts MedicalSchool, Worcester, Mass (SEA, JHG, TSF, GR).
Dr Majumdar is a Population Health Investigator supported by the Alberta HeritageFoundation for Medical Research and a New Investigator supported by the CanadianInstitutes of Health Research. Drs Majumdar and Gurwitz were corecipients of a ResearchMentorship Award from the Society of General Internal Medicine.
Address correspondence to: Jerry H. Gurwitz, MD, Executive Director, Meyers PrimaryCare Institute, 630 Plantation St, Worcester, MA 01605. E-mail: email@example.com and firstname.lastname@example.org.
Advance Data for Vital and Health Statistics: No. 337.
1. Cherry DK, Burt CW, Woodwell DA. National Ambulatory Medical Care Survey:2001 summary. Hyattsville, Md: National Center for Health Statistics; 2003.
Arch Intern Med.
2. Oliveria SA, Lapuerta P, McCarthy BD, L'Italien GJ, Berlowitz DR, Asch SM.Physician-related barriers to the effective management of uncontrolled hypertension. 2002;162:413-420.
Arch Intern Med.
3. Hyman D, Pavlik VN. Self-reported hypertension treatment practices among primarycare physicians. 2000;160:2281-2286.
4. Hajjar I, Kotchen TA. Trends in prevalence, awareness, treatment, and controlof hypertension in the United States, 1998-2000. . 2003;290:199-206.
5. National Committee for Quality Assurance. The Health Plan Employer Dataand Information Set (HEDIS®). Available at: http://www.ncqa.org/Programs/HEDIS.Accessed November 15, 2003.
Arch Intern Med.
6. Alexander M, Tekawa I, Hunkeler E, et al. Evaluating hypertension control in amanaged care setting. 1999;159:2673-2677.
Arch Intern Med.
7. Joint National Committee on Prevention, Detection, Evaluation and Treatmentof High Blood Pressure. The Sixth Report of the National Committee on Prevention,Detection, Evaluation, and Treatment of High Blood Pressure.1997;157:2413-2446.
Stata Cross-Sectional Time-Series Reference Manual Release 8.
8. StataCorp. College Station, Tex: Stata Press; 2003:132-140.
N Engl J Med
9. Berlowitz DR, Ash AS, Hickey EC, et al. Inadequate management of blood pressurein a hypertensive population. . 1998;339:1957-1963.
10. Knight EL, Bohn RL, Wang PS, Glynn RJ, Mogun H, Avorn J. Predictors ofuncontrolled hypertension in ambulatory patients. . 2001;38:809-814.
N Engl J Med
11. McGlynn EA, Asch SM, Adams J, et al. The quality of health care delivered toadults in the United States. . 2003;348:2635-2645.
N Engl J Med
12. Mechanic D, McAlpine DD, Rosenthal M. Are patients' office visits with physiciansgetting shorter? . 2001;344:198-204.
N Engl J Med
13. Berlowitz DR, Ash A, Moskowitz MA. Inadequate management of blood pressurein a hypertensive population [letter]. . 1999;340:1594-1595.
14. Chobanian AV, Bakris GL, Black HR, et al. The Seventh Report of the JointNational Committee on Prevention, Detection, Evaluation, and Treatment of HighBlood Pressure. The JNC 7 Report. . 2003;289:2560-2571.
Ann Intern Med
15. Phillips LS, Branch WT, Cook CB, et al. Clinical inertia. .2001;135:825-834.
16. Neal B, MacMahon S, Chapman N. Effects of ACE inhibitors, calcium antagonists,and other blood-pressure-lowering drugs. . 2000;356:1955-1964.