Review of evidence and explanations for suboptimal screening and treatment of dyslipidemia in women

JGIM

854

Blackwell Publishing Ltd

R EV I EW

Kim et al Suboptimal Screening and Treatment of Dyslipidemia in Women

Review of Evidence and Explanations for Suboptimal Screening and Treatment of Dyslipidemia in Women

A Conceptual Model

Catherine Kim MD MPH Timothy P Hofer MD MS Eve A Kerr MD MPH

Screening and treatment rates for dyslipidemia in populationsat high risk for cardiovascular disease (CVD) are inappropri-ately low and rates among women may be lower than amongmen We conducted a review of the literature for possible expla-nations of these observed gender differences and categorizedthe evidence in terms of a conceptual model that we describeFactors related to physiciansrsquo attitudes and knowledge thepatientrsquos priorities and characteristics and the health caresystems in which they interact are all likely to play impor-tant roles in determining screening rates but are not wellunderstood Research and interventions that simultaneouslyconsider the influence of patient clinician and health systemfactors and particularly research that focuses on modifiablemechanisms will help us understand the causes of theobserved gender differences and lead to improvements in cho-lesterol screening and management in high-risk women Forexample patient and physician preferences for lipid and otherCVD risk factor management have not been well studied par-ticularly in relation to other gender-specific screening issuescosts of therapy and by degree of CVD risk better understand-ing of how available health plan benefits interact with thesepreferences could lead to structural changes in benefits thatmight improve screening and treatment

KEY WORDS

cardiac disparity outpatient cholesterolwomen

J GEN INTERN MED 2003 18854ndash863

SUBOPTIMAL MANAGEMENT OF DYSLIPIDEMIA IN WOMEN

C

ardiovascular disease (CVD) is the leading causeof death and morbidity among women approximately

512904 women and 455871 men died from CVD in 1999

1

Womenrsquos cholesterol profile can be modified with statintherapy and the incidence of coronary events reduced

2ndash6

However screening and treatment rates for CVD risk fac-

tors in high-risk populations may be inappropriately lowparticularly in the outpatient setting

7ndash9

In order to identify articles that would distinguishscreening and treatment of women apart from men wesearched

MEDLINE

for publications fitting these criteriaEnglish language human subjects adults 19 years andolder and publication date during or after 1996 to allowdissemination and implementation of the ATPII guidelinesreleased in 1993 We added the terms (

gender

OR

women

)AND (

therapy

OR

measurement

OR

treatment

OR

manage-

ment

) AND (

coronary

OR

cardiovascular

OR

cardiac

) AND(

cholesterol

OR

lipid

OR

dyslipidemia

OR

hypercholester-

olemia

) which resulted in 4019 articles to further focusthe search on management of cholesterol as opposed topure epidemiological or mechanistic studies we added thesearch term (

undertreatment

OR

compliance

OR

utilization

OR

adherenc

e OR

bias

OR

disparity

OR

guidelines

) whichresulted in 541 articles We included articles that focusedon secondary prevention and outpatient management andreported gender comparisons in screening and treatmentWe excluded articles that did not have original data anddid not focus on cholesterol management resulting in 520articles Finally we reviewed the reference lists of thesearticles and included articles that were missed in the orig-inal search (

n

=

3) We abstracted the screening therapyand successful treatment percentages for men and womenand

P

values or confidence intervals when availableThe majority of studies that have examined screening

and treatment for dyslipidemia did not report analyses bygender or the degree of disparity between men and womenThe results are in Table 1 Multiple studies examining peo-ple with CVD indicate that men have cholesterol measuredmore often treated more aggressively and have lower low-density lipoprotein levels than women

710ndash14

The reasons for gender differences in management ofdyslipidemia in high-risk groups are unclear In patientswith CVD and CVD risk equivalents such as diabetescholesterol measurement and treatment goals are identicalfor men and women measurement of low-density lipopro-tein cholesterol (LDL-C) and a treatment goal of less than100 mgdL is recommended

15

It is possible that sub-optimal management of dyslipidemia in women occursthrough the same mechanisms that cause gender dis-parities in referrals for CVD diagnostic and therapeutic

Received from the Division of General Internal Medicine Univer-sity of Michigan (CK TPH EAK) and the Ann Arbor VA Centerfor Practice Management and Outcomes Research VA Ann ArborHealthcare System (TPH EAK) Ann Arbor Mich

Address correspondence and requests for reprints to Dr Cath-erine Kim 300 North Ingalls Building Room 7C13 Ann ArborMI 48109 (email cathkimumichedu)

JGIM

Volume 18 October 2003

855

procedures Although the literature documenting suchdisparities is extensive again the mechanisms remainvirtually unexplained and efforts to explain them havetended to focus on clinician factors

16ndash23

Elucidating themultiple factors that contribute to gender disparities mayhelp structure quality improvement interventions for bothmen and women

MODEL

Patient clinician and health system factors may allcontribute to suboptimal management of dyslipidemiaDrawing from the Health Belief model

24

the Landon et alhealth care organization model

25

and the Jaen et al com-peting demands model

26

we have developed a con-ceptual model (Fig 1) for how these factors interact Atheoretical model should have face validity provide mea-surable variables and enhance understanding beyondwhat would be expected from consideration of individualfactors affecting preventive service delivery In our concep-tual model we have incorporated the concept of patientperception of risk and subsequent behavior from the HealthBelief model the association between detailed structuralcharacteristics of health systems and physician behaviorfrom Landonrsquos model and the role of physician character-istics and the idea of competing interests from the Jaenet al model

The Health Belief model argues that health behaviorsare related to personal beliefs about susceptibility to dis-ease seriousness of disease benefit of intervention andrisk of intervention In this model individuals who do notbelieve they are at high risk of disease are unlikely to pur-sue preventive health behavior even if the benefits of thebehavior are high and the risks are low and individualswho believe they are at high risk of disease may pursuepreventive health behavior even if the benefits are low andthe risks are high

24

This model has proven useful in under-standing and predicting many preventive health behaviorsincluding diet and exercise However the model is not asuseful in examining the barriers to acting on such beliefsie barriers related to the structural organization of thehealth care system and barriers related to specific aspectsof the health care visit such as limited time In the HealthBelief model the barriers are limited primarily to thepatientrsquos perceived barriers to behavior change

In contrast the Landon et al model of health care organ-ization focuses on characteristics of the health care systemthat can influence health care delivery

25

In the Landon et almodel disease processes and outcomes can be influencedby financial incentives management strategies such as utili-zation review structure of care such as the location of thepractice site and staffing patterns and finally normativeinfluences such as the culture of the organization Thestrengths of this model are that it details health plan and

Table 1 Studies That Reported Results by Gender in Screening or Treatment for Dyslipidemia in Patients with Cardiovascular Disease

Data Source Results Extent of Disparity

Randomized trial56 Women less likely to receive lipid treatment and to be at goal than men

35 vs 55 (P lt 05) 29 vs 48 (P = 001)

Registry59 Women less likely to have lipid levels measured than men

Not given

Registry49 Women less likely to receive lipid treatment and to be at goal than men

44 vs 45 95 vs 161 (P value not given)

Registry1213 Women less likely to be at treatment goal than men

P lt 001 (percents not given)

Registry126 Women less likely to receive statins than men 82 vs 133 (P value not given)Pharmacy database127 Women less likely to receive statins than men 135 vs 208 (P lt 05)Patient survey60 Women less likely to receive lipid treatment and

to be at goal than men 298 vs 396 203 vs 219

(P value not given)Electronic medical record54 Women less likely to be screenedtreated

than men243 vs 373 (P lt 0001)

Chart review48 Women less likely to be treated than men 33 vs 48 (P = 047)Chart review95 Women less likely to be at lipid goal than men P lt 05 (percents not given)Chart review128 Women less likely to be screened treated

and at goal than men 35 vs 50 21 vs 31 23 vs 33

(P lt 0001)Chart review97 Women less likely to be screened and

at lipid goal 51 vs 68 (P = 001) 25 vs 34

(P = 043)Randomized trial58 No gender difference in screening or treatmentPatient survey129 No gender differences in screening or treatmentChart review57 No gender differences in treatmentChart review130 No gender difference in screening or treatmentPharmacy database131 No gender difference in treatmentPharmacy database132 No gender difference in treatmentPharmacy database133 No gender difference in treatmentPatient survey80 Women more likely to be treated 64 vs 81 (P lt 05)

856


JGIM

provider group characteristics that are probably influentialbut have not been the focus of extensive research Such amodel is extremely useful in conceptualizing changes tohealth care organizations in order to improve care but doesnot necessarily address the patientrsquos and providerrsquos per-ceptions of risk or barriers nor their interaction with thehealth system

Finally the Jaen et al health care model is posited on atheory of competing interests

26

In the Jaen et al model thepatient the physician and the practice environment are sep-arate domains that interact during the health care visit Themodel emphasizes the physicianrsquos role in delivering pre-ventive services specifically physicianrsquos skills and atti-tudes It also puts forth the idea of competing or alternativedemands for the physicianrsquos time as a physician barrierThis model is extremely useful for illustrating the physi-cianrsquos perceptions of barriers to provision of health servicesand is also valuable in that it empathizes with the clinicianand pinpoints a potentially reversible barrier rather thanplacing blame on the individual clinicianrsquos character As aresult of research showing that physician-level variation issmall compared with patient and health system variationwe believe that the physicianrsquos behavior is more heavilyinfluenced by the environment of the health system egvariable such as ldquolack of timerdquo may be more of a health systemcharacteristic than a physician-level characteristic and thatwomenrsquos agendas for screening play a more important role

Our model postulates that perceptions of the risks andbarriers to screening and treatment of CVD risk factors willaffect the clinicianrsquos behavior and the patientrsquos behaviorduring the health care visit We further hypothesize that

these perceptions can be partially predicted from patientcharacteristics such as gender In addition the health sys-tem structure affects screening and treatment of CVD riskfactors by affecting clinician behavior and patient behaviorWe use this model in framing the following review ofpatient clinician and health system variables that maycontribute to gender differences in management and willrefer back to it throughout the paper Although conceptu-alized for CVD risk factor management this model mayeasily be applied to understand gender differences in themanagement of other diseases as well

POSSIBLE PATIENT FACTORS

Multiple studies suggest that both mutable andimmutable patient-related factors such as access to carecompeting health issues and prioritization of these issuescommunication and decision-making preferences demo-graphics (eg race and age) and disease severity may playa role in gender disparities for hyperlipidemia treatment(Table 2)

Access to Care

Management of dyslipidemia depends on adequateaccess to clinicians and it is possible that women in certainhigh-risk groups have decreased access compared withmen Women

27

particularly uninsured women

28

citegreater cost barriers for access to care than men which inturn are associated with decreased preventive servicessuch as cholesterol screening In addition the quality ofthis care may be affected by cost barriers Quality care is

FIGURE 1 Conceptual model of patient clinician and health system factors affecting management of cholesterol in high-risk patients

JGIM


857

characterized by high continuity reasonable availabilityand good clinician communication and predicts use ofpreventive services in women

29

independent of a regularsource of care and insurance status

3031

Although theseanalyses did not stratify by coronary risk they did focuson a middle-aged population for examination of cholesterolscreening

Cost Barriers

Women who experience cost barriers for clinician accessmay also experience cost barriers for laboratory testingand medication copayment and decreased copayment hasbeen associated with improved performance of preventiveinterventions

32

These cost barriers may have been partiallyalleviated by Medicare or Medicaid Among Medicare bene-ficiaries women with known coronary disease may bemore likely to be on statin therapy than men

33

althoughin another analysis women reported greater difficulty thanmen in obtaining medical care and prescribed medicationsand women have reported delaying care owing to cost anddissatisfaction with the ease of getting to a physician moreoften than men

27

Medicaid patients had less frequentcholesterol screening than those with private insurancebut Medicaid patients did not cite specific barriers as aresult of Medicaid suggesting that the poorer Medicaidpopulation also faced cost barriers other than their insur-ance status

34

Not surprisingly the few studies examining

gender socioeconomic status and cholesterol levels in theUnited States suggest that lower socioeconomic status isassociated with adverse lipid levels to a greater extent inwomen than in men

1435

Competing Health Issues and CommunicationDecision-Making Styles

Direct observation of primary care visits has demon-strated that women receive gender-specific screeningspecifically mammography more often than cholesterolscreening

3637

Since these studies have not examinedwomen at high risk for CVD separately it is possible thathigh-risk women are managed differently than low-riskwomen However across all age groups and ethnicities womenare more concerned about breast cancer than coronarydisease

38

which might lead to their initiating discussionof this topic Since the issues addressed in the outpatientvisit are largely driven by patient concerns and needs

3940

andthe length of the visit is fixed this competing concern coulddecrease the attention paid to cholesterol management

26

Time constraints might be further compounded by womenrsquosdifferent communication and decision-making style

4142

which tends to be associated with longer office visits

4344

Demographics

Differences in lipid management between men andwomen may partially reflect a disparity in other factors

Table 2 Possible Mechanisms for Gender Disparities in Patients with Cardiovascular Disease

References

Possible patient factorsDecreased access to care among uninsured women 28 29Cost barriers greater in women 27 34Lower socioeconomic status 14 35Womenrsquos prioritization of cholesterol below gender-specific screening 36 37Womenrsquos perception of CVD risk compared with risk of other diseases 38Womenrsquos different communication and decision-making preferences 41 42 44Womenrsquos increased age 48 132African-American race 8 50Increased comorbidity 1 56Womenrsquos lower rate of revascularization 48 58Substitution of hormone replacement therapy for lipid-specific therapy 61

Possible provider factorsPerception of CVD risk inappropriately low treatment threshold in women inappropriately high 49 60 77 78Perception of CVD risk higher for revascularized patients but women undergo revascularization less 22 58Cardiologists specialty protective but women may see them less frequently 80 81 84 85Providers prioritize cholesterol management below other gender-specific screening 87Providers overestimate the amount of care they provide for women more than men 108Women physicianrsquos different communication and decision-making preferences 43Younger physicians with more recent training more likely to enforce prevention less likely to see women 93 94

Possible system factorsCase-management programs and lipid clinics may reduce gender disparities 97 99Cardiac rehabilitation programs underused by women 100 134Measurement factors 107Other health system factors such as profit status model type referral management ndash

This factor has already been demonstrated to differ between men and women with cardiovascular disease (CVD) for dyslipidemia screeningtreatment or goals

858

Kim et al Suboptimal Screening and Treatment of Dyslipidemia in Women JGIM

such as age Multiple studies have documented that elderlypatients are not treated as aggressively as younger patientsafter a myocardial infarction45ndash47 despite recommendationsfor management of dyslipidemia extend into older agegroups15 However age does not explain the entire dis-parity as gender discrepancies in lipid management persistin studies of older patients4849 Similarly a greater pro-portion of women with dyslipidemia are from minoritypopulations and gender differences may represent racialdifferences to some extent5051 However managed careregistries have documented gender differences in lipidmanagement that persist after adjustment for race genderdifferences may be greater in minority populations than inwhite populations52ndash54 Also difficulties with access to caretend to be greater in African-American women comparedwith African-American men and white women comparedwith white men27 Examination of US population-baseddata indicates that gender race and socioeconomic statusprobably exert independent effects on lipid levels143555

Disease Severity

Since women with CVD tend to have a greater numberof comorbid conditions than men with CVD1 comorbiditymay partially explain differences in cholesterol manage-ment56 Despite the effectiveness of therapy in women over-all5 in actual practice sicker patients may be prescribedtherapy less often or be able to comply with therapy lessoften than healthier patients Given the excellent clinicalrisk models that currently exist for CVD adjustment fordisease severity differences is possible but may requireinformation that is not always available and therefore itis not always performed

The effect of comorbidity upon cholesterol manage-ment is complex and probably goes beyond the number ofcomorbid conditions Analyses of dyslipidemic patientsadmitted for myocardial infarction suggest that patientswith no comorbidity or severe comorbidity may be treatedless often than patients with moderate comorbidity57 Inaddition specific comorbid conditions such as hyperten-sion may increase awareness of CVD After myocardialinfarction patients with hypertension tend to be treatedmore often for hypercholesterolemia than patients withouthypertension33 despite evidence of treatment for cholesterol-lowering therapy regardless of blood pressure status in thisgroup of patients A similar association between presenceof hypertension and cholesterol screening exists in abroader population30 suggesting that these 2 CVD riskfactors are probably linked cognitively or in a more sys-tematic fashion ie through prompts or guidelines

Similarly procedures such as revascularizationincrease the likelihood of appropriate lipid therapy4858

This could reflect increased recognition by clinicians ofCVD status better underlying health of patients whounderwent revascularization compared with those who didnot the increased likeliness of these patients to comply withtherapy or other factors associated with revascularization

Of note women veterans with CVD were still less likely thanmale veterans with CVD to have their cholesterol measuredeven after adjustment for age coronary procedures anginaand other CVD risk factors59

Other Confounders

Other confounders for gender probably exist In 2 stud-ies that also adjusted for other CVD risk factors body massindex years of education current employment and cardi-ology visit female gender no longer predicted underuse oflipid therapy in women with CVD3360 In addition womenmay be treated less often for hypercholesterolemia thanmen because of the substitution of hormone replacementtherapy (HRT) for specific lipid-lowering therapy since HRTis associated with lowered LDL-C levels61 Unfortunatelyhormone replacement in and of itself cannot lower LDL-Cto goal in women with established coronary disease62 andpreviously held beliefs about the cardiovascular indicationsfor HRT have been discredited6364 Finally women tend tobe affected by disorders that are not always included inrisk-adjustment models but that nonetheless may affectmanagement of CVD risk factors Specifically womensuffer from higher rates of disability65 obesity66 andanxiety and affective disorders67 that have been demon-strated to adversely affect health services such as cancerscreening6869 and work-up of pain70 although the associ-ation between these disorders and gender differences in themanagement of CVD has not been studied

POSSIBLE CLINICIAN FACTORS

It is logical that individual clinician practices influencescreening and treatment of cholesterol71 and contribute todifferences in hyperlipidemia management between menand women at high risk for CVD Studies that account forpatient case-mix and clustering of patients have concludedthat variation between individual clinician practicesaccounts for less than 5 of variation in practice72ndash76

Nonetheless if clinicians on the whole are treating womenless aggressively than men then clinician factors could beimportant Such clinician factors include perception of thepatientrsquos CVD risks and the benefits of treatment con-fidence in the ability to manage cholesterol disorders pri-oritization of other preventive services over CVD risk factormanagement and communication and decision-makingstyles (Table 2)

Misperceptions of Risk

Gender differences in lipid management could reflectdifferent clinician treatment thresholds for men and womenthat are dictated by factors other than coronary disease49

In turn these thresholds may be affected by perceptionsof coronary risk that are inappropriately low for womenPerception of risk could be affected by the lower prevalenceof coronary disease in women compared with men althoughcoronary disease is still common in older women7778 It

JGIM Volume 18 October 2003 859

could also be affected by other factors that increased thesalience of coronary disease to the clinician such asrevascularization58 since men undergo revascularizationat higher rates than women162279 this factor could increaseclinician awareness of CVD in men more often than in women

Perception of risk could also explain why specialistswith heightened awareness of coronary disease would bemore likely to treat dyslipidemia8081 Specialty training mayalso reflect ability or confidence in onersquos ability to treat8283

and the degree to which cholesterol testing is included inthe particular specialistrsquos role in the patientrsquos healthcare A significant number of women receive care from ageneralist physician only and therefore may overall beless likely to receive preventive testing8485 Visits toobstetrician-gynecologists may actually lead to increasedcholesterol screening although this finding may be pri-marily in populations of low-risk women85 has not beenconsistently documented86 and subsequent treatmentrates were not examined

Prioritization

Physicians along with their female patients mayprioritize gender-specific screening ahead of cholesterolmanagement When presented with a vignette presenting53-year-old woman clinicians ranked cancer screeningahead of cholesterol testing in importance87 This may bebecause of misperceptions about the risk of coronary dis-ease in relation to breast cancer or driven by other providerconcerns such as liability Missed cases of breast cancerare the most common cause of litigation in the UnitedStates8889 and this concern may drive certain cliniciansto focus on the breast examination and mammography dis-cussions at the expense of other health issues

Communication and Decision-Making Styles

Women patients tend to prefer women physicians whomay also prioritize gender-specific screening over choles-terol screening although studies to date have demon-strated that women physicians generally perform manypreventive services at comparable rates or more often thanmale physicians8590ndash92 High-risk populations for CVD werenot examined separately Women physicians also tend tohave more participatory and social communication stylesthan male physicians which can be associated with longervisit length and contribute to time constraints4143 Youngerphysicians93 and physicians with fewer years in training94

tend to provide increased preventive services includingcholesterol screening but it is not known whether thesephysicians see women less often or provide different careto women Finally individual clinicians may attract differentpatient populations and tailor or impose their unique practicestyles upon that population95

POSSIBLE HEALTH SYSTEM FACTORS

Although variation attributable to the patient tends toaccount for the majority of variation in treatment the

health system as represented by the facility can accountfor variation in practice as well76 As a result in part ofthe large numbers of facilities needed for an adequatelypowered analysis specific system-level factors have notbeen examined for their effect on gender differences inscreening or treatment for dyslipidemia (Table 2) There issome evidence that the presence of disease managementprograms can significantly influence screening andtreatment rates Case management96 lipid clinics97 andmultidisciplinary CVD programs in general98 have beeneffective in decreasing lipid levels in patients with knowncoronary disease LaBresh et al and Bramlet et al foundthat men with CVD were more likely to respond to lipid-lowering therapy than women with CVD with standardcare however gender differences were absent in patientswho were referred to nurse management9799 Women mayalso have lower participation rates in cardiac rehabilitationprograms after myocardial infarction100 but it is not clearhow much of this is because of patient preferences101 orgender biases in referral practices102 To our knowledgeother gender differences in the associations between otherhealth system factors (Fig 1) and lipid management strate-gies have not been reported

METHODS OF MEASUREMENT

The severity of the problem depends partially on themethod of measurement103 The use of medical records andclaims data may not adequately record services providedparticularly discussion of issues surrounding cholesterolscreening and management103ndash106 Quality measurementstudies that have trained experienced actors to serve asstandardized patients for several common conditions havecaptured a greater number of services provided in thevisit94107 These studies found that medical record abstrac-tion underestimated compliance with preventive measuresby as much as 26 and that patients did not recall asignificant portion of what they had been told during thevisit94107 On the other hand surveys of providers tendto overestimate provider compliance with cholesterolguidelines108

The instrument used to measure treatment may alsoaffect the estimates of the relative importance of patientclinician and health system factors from analyses of largerdatabases Larger amounts of variation in clinician practiceis seen when processes of care that are linked to interme-diate outcomes are examined instead of outcomes aloneFor example measurement of lipid level alone demon-strates that patient and health system factors are theprimary determinants of lipid level However clinician var-iability in practice is more pronounced for an indicator thatmeasures whether a statin was prescribed for an elevatedlipid level76 Therefore the extent of the problem may varydepending on the source and construction of informationused Whether or not the method of measurement affectslipid management in men and women differently has notbeen examined

860 Kim et al Suboptimal Screening and Treatment of Dyslipidemia in Women JGIM

LIMITATIONS OF SOLUTIONS FOR SUBOPTIMAL SCREENING AND TREATMENT

To date proposed solutions to improve managementof CVD risk factors for both men and women have includededucating clinicians and increasing clinician awarenessand accountability through feedback reports71 Unfortunatelyprofiling for resource utilization and clinician ldquoaccountabilityrdquopurposes has not been shown to affect prescribing behavioror lab test ordering109ndash111 perhaps because these ldquoreportcardsrdquo do not accurately reflect a clinicianrsquos case-mix andbecause of limited power to detect differences among cli-nicians75 Also clinicians tend not to think of screening andtreatment failures as clinician-based problems71

Health system level interventions such as case-management computerized reminders and patient edu-cation programs have demonstrated success in improvingprocess or outcomes in high-risk populations9698112ndash116

but the patient population practice setting and programsvary widely across studies In a meta-analysis of interven-tions to improve CVD risk factor management programsthat targeted several levels of care including the structuralorganization of care and patient education tended to bethe most successful117 To our knowledge only the pre-viously mentioned studies by LaBresh et al and Bramletet al have compared the effects of these programs on lipidlevels between women and men9799

CONCLUSIONS

Despite the availability of effective medical therapiesto reduce CVD mortality current literature suggests thatwomen with CVD experience suboptimal cholesterol man-agement The reasons underlying the gender difference andpoor management overall are not well understood Thereforeit is unclear how to reduce such gender disparities andthese disparities may translate to significantly higher ratesof CVD events and mortality for women The gap betweenresearch and actual practice has led Healthy People 2010to support the study of the management of dyslipidemia118

and the National Heart Lung and Blood Institute to declarethe study of the translation of research results into practiceas part of its strategic plan for 2002ndash06119

Further research on disparities in lipid mismanagementshould focus on modifiable mechanisms Womenrsquos andmenrsquos preferences for lipid and other CVD risk factor man-agement have not been well studied particularly in relationto other gender-specific screening issues cost of therapyand by degree of CVD risk Understanding clinician prior-itization of cholesterol screening and management andgender-specific thresholds in management could providefurther insight into ldquoclinical inertiardquo71 Better understandingof how the structure of health care organizations partic-ularly specialty referral utilization management andpayment arrangements affect screening and treatment inwomen and men separately might also provide insight intodifferences in management For example understanding ofhow available health plan benefits interact with patient and

physician preferences for cholesterol management couldlead to structural changes in benefits that might improvescreening and treatment In general we found in ourreview that there is weak and often inconsistent evidencefor the importance of a wide variety of variables throughoutthe major domains of our conceptual model Yet there areno studies that consider more than a few variables ordomains in any single analysis What is most critical to thisresearch agenda is that patient clinician and health sys-tem defects be considered simultaneously in order to clarifywhich factors are most influential and modifiable

There are a number of reasons to pursue this researchagenda Investigation of gender disparities in CVD risk andlipid management may shed light on gender disparities inother disease areas The area of lipid management has awell-developed evidence base supporting a set of widelyaccepted and specific guidelines thus reducing reasonablevariations in practice The presence of information thatenables accurate assessment of CVD risk in men andwomen can reduce concern about confounding by diseaseseverity Finally it seems likely that insights about possiblemechanisms of disparities outlined in our model for CVDmay be generalizable to other diseases120ndash125 particularlythose managed in the outpatient setting

Although clinicians may not be able to single-handedlychange adherence patterns they can be aware of issues ofscreening and treatment during the health care visit Whenmanaging a woman at high risk for CVD clinicians shouldrespect the patientrsquos agenda but also attempt to negotiatethat agenda so that interventions such as screening andtreatment of cholesterol occur The time for such negotia-tion can occur by delegating discussions to ancillary staffor automating testing procedures decreasing the amountof time spent on other screening recommendations forwhich the patient is at lower risk or having the patientreturn for another visit Clinicians need to be aware of theservices their health system or insurance plan offers to helpmanage dyslipidemia in the face of competing time con-straints such as wellness clinics preventive cardiologyservices nutritional counseling exercise programs casemanagement programs and social workers who can edu-cate patients about their eligibility for health care benefitsFinally they should be sympathetic to the barriers thatwomen particularly those of lower socioeconomic statusface in successfully implementing such goals

Dr Kim is supported by an American Diabetes Association JuniorFaculty Award Dr Kerr is supported by an Advanced ResearchCareer Development Award from the Department of VeteransAffairs Health Services Research and Development Service DrHofer is supported by grant 1P20HS011540-01 from the Agencyfor Health Research and Quality

REFERENCES1 American Heart Association American Heart Association 2002

Heart and Stroke Statistical Update Dallas TX American HeartAssociation 20011ndash38


2 Miettinen T Pyorala K Olsson A et al Cholesterol-lowering therapyin women and elderly patients with myocardial infarction orangina pectoris Circulation 1997964211ndash8

3 McPherson R Genest J Angus C Murray P The WomenrsquosAtorvastatin Trial on Cholesterol (WATCH) frequency of achievingNCEP-II target LDL-C levels in women with and without estab-lished CVD Am Heart J 2001141949ndash56

4 LIPID Study Group Prevention of cardiovascular events and deathwith pravastatin in patients with coronary heart disease and a broadrange of initial cholesterol levels N Engl J Med 19983391349ndash57

5 Lewis S Sacks F Mitchell J et al Effect of pravastatin on car-diovascular events in women after myocardial infarction TheCholesterol and Recurrent Events (CARE) trial J Am Coll Cardiol199832140ndash6

6 Waters D Higginson L Gladstone P Boccuzzi S Cook T Lesperance JEffects of cholesterol lowering on the progression of coronaryatherosclerosis in women a Canadian Coronary AtherosclerosisIntervention Trial (CCAIT) substudy Circulation 1995922404ndash10

7 Meigs J Stafford R Cardiovascular disease prevention practicesby US physicians for patients with diabetes J Gen Intern Med200015220ndash8

8 Qureshi A Suri M Guterman L Hopkins L Ineffective secondaryprevention in survivors of cardiovascular events in the US popu-lation Report from the Third National Health and Nutrition Exam-ination Survey Arch Intern Med 20011611621ndash8

9 Saadine J Engelgau M Beckles G Gregg E Thompson T Venkat-Narayan K A diabetes report card for the United States qualityof care in the 1990s Ann Intern Med 2002136565ndash74

10 Brown D Giles W Greenlund K Croft J Disparities in cholesterolscreening falling short of a national health objective Prev Med200133517ndash22

11 Davis K Cogswell M Lee S Rothenberg R Koplan J Lipidscreening in a managed care population Public Health Rep1998113346ndash51

12 Pearson T The undertreatment of LDL-cholesterol addressing thechallenge Int J Cardiol 2000S23ndash28

13 Pearson T Laurora I Chu H Kafonek S The Lipid TreatmentAssessment Project (L-TAP) Arch Intern Med 2000160459ndash67

14 Gardner C Winkleby M Fortmann S Population frequencydistribution of non-high-density lipoprotein cholesterol (ThirdNational Health and Nutrition Examination Survey 1988ndash94) AmJ Cardiol 200086299ndash304

15 Expert Panel on Detection Evaluation and Treatment of HighBlood Cholesterol in Adults Executive summary of the thirdreport of the National Cholesterol Education Program (NCEP)Expert Panel on Detection Evaluation and Treatment of HighBlood Cholesterol in Adults (Adult Treatment Panel III) JAMA20012852486ndash97

16 Ayanian J Epstein A Differences in the use of procedures betweenwomen and men hospitalized for coronary heart disease N EnglJ Med 1991325221

17 Weintraub W Kosinski A Wenger N Is there a bias againstperforming coronary revascularization in women Am J Cardiol1996781154ndash60

18 Schulman K Berlin J Harless W et al The effect of race and sexon physiciansrsquo recommendations for cardiac catheterization NEngl J Med 1999340618ndash26

19 Roger V Farkouh M Weston S et al Sex differences in evaluationand outcome of unstable angina JAMA 2000283646ndash52

20 Rathore S Chen J Wang Y Radford M Vaccarino V KrumholzH Sex differences in cardiac catheterization the role of physiciangender JAMA 20012862849ndash56

21 Alter D Naylor C Austin P Tu J Biology or bias practice patternsand long-term outcomes for men and women with acute myocar-dial infarction J Am Coll Cardiol 2002391909ndash16

22 Ghali W Faris P Galbraith P et al Sex differences in access tocoronary revascularization after cardiac catheterization impor-tance of detailed clinical data Ann Intern Med 2002136723ndash32

23 Shaw L Miller D Romeis J et al Gender differences in the non-

invasive evaluation and management of patients with suspectedcoronary artery disease Ann Intern Med 1991120559

24 Janz N Becker M The Health Belief Model a decade later HealthEduc Q 1984111ndash47

25 Landon B Wilson I Cleary P A conceptual model of the effectsof health care organizations on the quality of medical care JAMA19982791377ndash82

26 Jaen C Stange K Nutting P Competing demands of primary carea model for the delivery of clinical preventive services J Fam Pract199438166ndash71

27 Janes G Blackman D Bolen J et al Surveillance for use ofpreventive health-care services by older adults 1995ndash97 MorbMortal Wkly Rep CDC Surveill Summ 19994851ndash88

28 Ayanian J Weissman J Schneider E Ginsburg J Zaslavsky AUnmet health needs of uninsured adults in the United StatesJAMA 19982842061ndash9

29 Bindman A Grumbach K Osmond D Vranizan K Stewart APrimary care and receipt of preventive services J Gen Intern Med199611269ndash76

30 Corbie-Smith G Flagg E Doyle J OrsquoBrien M Influence of usualsource of care on differences by raceethnicity in receipt of preventiveservices J Gen Intern Med 200217458ndash64

31 Mainous A Hueston W Love M Griffith C Access to care for theuninsured is access to a physician enough Am J Public Health199989910ndash2

32 Lurie N Manning W Peterson C Goldberg G Phelps C Lillard LPreventive care do we practice what we preach Am J PublicHealth 198777801ndash4

33 Ayanian J Landon B Landrum M Grana J McNeil B Use ofcholesterol-lowering therapy and related beliefs among middle-agedadults after myocardial infarction J Gen Intern Med 20021795ndash7

34 Hueston W Spencer E Kuehn R Differences in the frequency ofcholesterol screening in patients with Medicaid compared withprivate insurance Arch Fam Med 19954331ndash4

35 Luepker R Rosamond W Murphy R et al Socioeconomic statusand coronary heart disease risk factor trends The MinnesotaHeart Survey Circulation 1993882172ndash9

36 Cooper G Goodwin M Stange K The delivery of preventive servicesfor patient symptoms Am J Prev Med 200121177ndash81

37 Stange K Flocke S Goodwin M Kelly R Zyzanski S Direct obser-vation of rates of preventive service delivery in community familypractice Prev Med 200031167ndash76

38 Mosca L Jones W King K Ouyang P Redberg R Hill M Awarenessperception and knowledge of heart disease risk and preventionamong women in the United States American Heart AssociationWomenrsquos Heart Disease and Stroke Campaign Task Force ArchFam Med 20009506ndash15

39 Marvel M Epstein R Flowers K Beckman H Soliciting thepatientrsquos agenda have we improved JAMA 1999281283ndash7

40 Kaplan C Siegel B Madill J Epstein A Communication and themedical interview strategies for learning and teaching J GenIntern Med 1997 S49ndash55

41 Hall J Roter D Patient gender and communication with physiciansresults of a community-based study Womens Health 1995177ndash95

42 Elderkin-Thompson V Waitzkin H Differences in clinical commu-nication by gender J Gen Intern Med 199914112ndash21

43 Roter D Hall J Aoki Y Physician gender effects in medicalcommunication a meta-analytic review JAMA 288756ndash64

44 Kaplan S Gandek B Greenfield S Rogers W Ware J Patient andvisit characteristics related to physiciansrsquo participatory decision-making style Med Care 1995331176ndash87

45 Ayanian J Landrum M McNeil B Use of cholesterol-loweringtherapy by elderly adults after myocardial infarction Arch InternMed 20021621013ndash9

46 Lemaitre R Furberg C Newman A et al Time trends in the useof cholesterol-lowering agents in older adults the CardiovascularHealth Study Arch Intern Med 19981581761ndash8

47 Aronow W Underutilization of lipid-lowering drugs in older per-sons with prior myocardial infarction and a serum low-density


lipoprotein cholesterol gt 125 mgdl Am J Cardiol 199882668ndash9A6A8

48 Di Cecco R Patel U Upshur R Is there a clinically significant genderbias in post-myocardial pharmacological management in the older (gt60)population of a primary care practice BMC Fam Pract 200238

49 Laubach E Otto C Schwandt P Toward better therapy of hyper-cholesterolemia Arch Intern Med 20001602685ndash6

50 Harris M Racial and ethnic differences in health care access andhealth outcomes for adults with type 2 diabetes Diabetes Care200124454ndash9

51 Nelson K Norris K Mangione C Disparities in the diagnosis andpharmacologic treatment of high serum cholesterol by race andethnicity Arch Intern Med 2002162929ndash35

52 Wisdom K Fryzek J Havstad S Anderson R Dreiling M Tilley BComparison of laboratory test frequency and test results betweenAfrican-Americans and Caucasians with diabetes opportunity forimprovement findings from a large urban health maintenanceorganization Diabetes Care 199720971ndash7

53 Cook C Erdman D Ryan G et al The pattern of dyslipidemiaamong urban African-Americans with type 2 diabetes DiabetesCare 200023319ndash24

54 Maviglia S Teich J Fiskio J Bates D Using an electronic medicalrecord to identify opportunities to improve compliance with cho-lesterol guidelines J Gen Intern Med 200116531ndash7

55 Winkleby M Cubbin C Ahn D Kraemer H Pathways by whichSES and ethnicity influence cardiovascular risk factors Ann N YAcad Sci 1999896191ndash209

56 Miller M Byington R Hunninghake D Pitt B Furberg C Sex biasand underutilization of lipid-lowering therapy in patients withcoronary artery disease at academic medical centers in the UnitedStates and Canada Arch Intern Med 2000160343ndash7

57 Majumdar S Gurwitz J Soumerai S Undertreatment of hyperli-pidemia in the secondary prevention of coronary artery diseaseJ Gen Intern Med 199914711ndash7

58 McBride P Schrott H Plane M Underbakke G Brown R Primarycare practice adherence to National Cholesterol Program guide-lines for patients with coronary heart disease Arch Intern Med19981581238ndash44

59 Sloan K Sales A Willems J et al Frequency of serum low-densitylipoprotein cholesterol measurement and frequency of results lt100 mgdl among patients who had coronary events (NorthwestVA Network Study) Am J Cardiol 2001881143ndash6

60 Vanuzzo D Pilotto L Ambrosio G et al Potential for cholesterol low-ering in secondary prevention of coronary heart disease in Europefindings from EUROASPIRE study Atherosclerosis 2000153505ndash17

62 The Writing Group for the PEPI Trial Effects of estrogen or estro-genprogestin regimens on heart disease risk factors in postmen-opausal women JAMA 1995273199ndash208

63 Hulley S Grady D Bush T et al Randomized trial of estrogenplus progestin for secondary prevention of coronary heart diseasein postmenopausal women Heart EstrogenProgestin Replace-ment Study (HERS) Res Group JAMA 1998280605ndash13

64 Womens Health Initiative Primary prevention with estrogenpro-gestin JAMA 2002

65 Iezzoni L McCarthy E Davis R Siebens H Mobility difficulties arenot only a problem of old age J Gen Intern Med 200116235ndash43

66 McTigue K Garrett J Popkin B The natural history of the devel-opment of obesity in a cohort of young US adults between 1981and 1998 Ann Intern Med 2002136857ndash64

67 Kessler R McGonagle K Zhao S et al Lifetime and 12-monthprevalence of DSM-III-R psychiatric disorders in the United StatesResults from the National Comorbidity Survey Arch Gen Psychi-atry 1994518ndash19

68 Iezzoni L McCarthy E Davis R Siebens H Mobility impairmentsand use of screening and preventive services Am J Public Health200090955ndash61

69 Wee C McCarthy E Davis R Phillips R Screening for cervical andbreast cancer is obesity an unrecognized barrier to preventivecare Ann Intern Med 2000132697ndash704

70 Garber M Bergus G Dawson J Wood G Levy B Levin I Effectof a patientrsquos psychiatric history on physiciansrsquo estimation of prob-ability of disease J Gen Intern Med 200015204ndash6

71 Phillips L Branch W Cook C et al Clinical inertia Ann InternMed 2001135824ndash34

72 Greenfield S Kaplan S Kahn R Ninomiya J Griffith J Profilingcare by different groups of physicians effects of patient case-mix(bias) and physician-level clustering on quality assessmentresults Ann Intern Med 2002136111ndash21

73 Orav E Wright E Palmer R Hargraves J Issues of variability andbias affecting multisite measurement of quality of care Med Care1996S87ndash101

74 Sixma H Spreeuwenberg P van der Pasch M Patient satisfactionwith the general practitioner a two-level analysis Med Care199836212ndash29

75 Hofer T Hayward R Greenfield S Wagner E Kaplan S Manning WThe unreliability of individual physician ldquoreport cardsrdquo for assessingthe costs and quality of care of a chronic disease JAMA19992812098ndash105

76 Krein S Hofer T Kerr E Hayward R Whom should we profileExamining diabetes care practice variation among primary careproviders provider groups and healthcare facilities Health ServRes 2002271159ndash80

77 Grover S Lowensteyn I Esrey K et al Do doctors accurately assesscoronary risk in their patients Preliminary results of the coronaryhealth assessment study BMJ 1995310975ndash8

78 Birdwell B Herbers J Kroenke K Evaluating chest pain ArchIntern Med 19931531991ndash5

79 Roger V Jacobsen S Weston S et al Sex differences in evaluationand outcome after stress testing Mayo Clin Proc 200277638ndash45

80 Ayanian J Landrum M Guadagnoli E Gaccione P Specialty ofambulatory care physicians and mortality among elderly patientsafter myocardial infarction N Engl J Med 20023471678ndash86

81 Stafford R Blumenthal D Specialty differences in cardiovasculardisease prevention practices J Am Coll Cardiol 1998321238ndash43

82 Hyman D Maibach W Flora J Fortmann S Cholesterol treatmentpractices of primary care physicians Public Health Rep1992107441ndash8

83 Marcelino J Feingold K Inadequate treatment with HMG-CoAreductase inhibitors by health care providers Am J Med1996100605ndash10

84 Wyn R Brown E Yu H Womenrsquos Use of Preventive Services TheCommonwealth Fund Survey Baltimore MD Johns HopkinsUniversity Press 1996

85 Henderson J Weisman C Grason H Are two doctors better thanone Womenrsquos physician use and appropriate care WomensHealth Issues 200212138ndash49

86 Giles W Anda R Jones D Serdula M Merritt R DeStefano FRecent trends in the identification and treatment of high bloodcholesterol by physicians Progress and missed opportunitiesJAMA 19932691133ndash8

87 Stange K Fedirko T Zyzanski S Jaen C How do family physiciansprioritize delivery of multiple preventive services J Fam Pract199438231ndash7

88 Osuch J Bonham V Morris L Primary care guide to managing abreast mass step-by-step work-up Medscape Womens Health199834

89 Barratt A Cockburn J Furnival C McBride A Mallon L Perceivedsensitivity of mammographic screening womenrsquos views on testaccuracy and financial compensation for missed cancers J Epi-demiol Community Health 199953716ndash20

90 Henderson J Weisman C Physician gender effects on preventivescreening and counseling an analysis of male and female patientsrsquohealth care experiences Med Care 2001391281ndash92

91 Cassard S Weisman C Plichta S Johnson T Physician gender andwomenrsquos preventive services J Womens Health 19976199ndash207

92 Franks P Clancy C Physician gender bias in clinical decision mak-ing screening for cancer in primary care Med Care 199331213ndash8


93 Schwartz J Lewis C Clancy C Kinosian M Radany M Koplan JInternistsrsquo practices in health promotion and disease preventionA Survey Ann Intern Med 199111446ndash53

94 Dresselhaus T Peabody J Lee M Wang M Luck J Measuringcompliance with preventive care guidelines J Gen Intern Med200015782ndash8

95 Harnick D Cohen J Schechter C Fuster V Smith D Effects ofpractice setting on quality of lipid-lowering management in patientswith coronary artery disease Am J Cardiol 1998811416ndash20

96 DeBusk R Miller N Superko H et al Case-management systemfor coronary risk factor modification after acute myocardialinfarction Ann Intern Med 1994120721ndash9

97 Bramlet D King H Young L Witt J Stoukides C Kaul A Managementof hypercholesterolemia practice patterns for primary care pro-viders and cardiologists Am J Cardiol 1997 39Hndash44H

98 McAlister F Lawson F Teo K Armstrong P Randomised trials ofsecondary prevention programs in coronary heart disease system-atic review BMJ 2001323957ndash62

99 LaBresh K Owen P Alteri C et al Secondary prevention in a car-diology group practice and hospital setting after a heart-careinitiative Am J Cardiol 20008523Andash29A

100 Evenson K Rosamond W Luepker R Predictors of outpatientcardiac rehabilitation utilization the Minnesota Heart SurgeryRegistry J Cardiopulm Rehabil 199818192ndash8

101 Blackburn G Foody J Sprecher D Park E Apperson-Hansen CPashkow F Cardiac rehabilitation participation patterns in a largetertiary care center evidence for selection bias J CardiopulmRehabil 200020189ndash95

102 Mosca L Han R Filip J Barriers for physicians to refer to cardiacrehabilitation and impact of a critical care pathway on rates ofparticipation Circulation 1998Indash811 Abstract

103 Stange K Zyzanski S Smith T et al How valid are medical recordsand patient questionnaires for physician profiling and healthservices research A comparison with direct observation ofpatients visits Med Care 199836851ndash67

104 Kerr E Krein S Vijan S Hofer T Hayward R Avoiding pitfalls inchronic disease quality management a case for the next generationof technical quality measures Am J Manag Care 200171033ndash43

105 Luck J Peabody J Dresselhaus T Lee M Glassman P How welldoes chart abstraction measure quality A prospective comparisonof standardized patients with the medical record Am J Med2000108642ndash9

106 Bloom S Harris J Thompson B Ahmed F Thompson J Trackingclinical preventive service use a comparison of the Health PlanEmployer Data and Information Set with the Behavioral RiskFactor Surveillance System Med Care 200038187ndash94

107 Peabody J Luck J Glassman P Dresselhaus T Lee M Com-parison of vignettes standardized patients and chart abstractiona prospective validation study of 3 methods for measuring qualityJAMA 20002831715ndash22

108 Headrick L Speroff T Pelecanos H Cebul R Efforts to improvecompliance with the National Cholesterol Education Programguidelines Results of a randomized controlled trial Arch InternMed 19921522490ndash6

109 Schechtman J Kanwal N Schroth W Elinsky E The effect of aneducation and feedback intervention on group-model andnetwork-model health maintenance organization physicianprescribing behavior Med Care 199533139ndash44

110 Mainous A Hueston W Love M Evans M Finger R An evaluationof statewide strategies to reduce antibiotic overuse Fam Med20003222ndash9

111 Balas E Boren S Brown G Ewigman B Mitchell J Perkoff GEffect of physician profiling on utilization meta-analysis of rand-omized clinical trials J Gen Intern Med 199611584ndash90

112 Renders C Valk G Franse L Schellveis F Van Eijk J van derWal G Long-term effectiveness of a quality improvement programfor patients with type 2 diabetes in general practice Diabetes Care2001241365ndash70

113 Baker A Lafata J Ward R Whitehouse F Divine G A web-baseddiabetes care management support system Jt Comm J QualImprov 200127179ndash90

114 Peters A Davidson M Application of a diabetes managed careprogram The feasibility of using nurses and a computer systemto provide effective care Diabetes Care 1998211037ndash43

115 Rubin R Kietrich K Hawk A Clinical and economic impact ofimplementing a comprehensive diabetes management program inmanaged care J Clin Endocrinol Metab 1998832635ndash42

116 Domurat E Diabetes managed care and clinical outcomes theHarbor City California Kaiser Permanente Diabetes Care SystemAm J Manag Care 199951299ndash307

117 Renders C Valk G Griffin S Wagner E Eijk van J Assendelft WInterventions to improve the management of diabetes in primarycare outpatient and community settings a systematic reviewDiabetes Care 2001241821ndash33

118 Centers for Disease Control and Prevention and Health HeartDisease and Stroke Healthy People 2010-Conference Edition1999 Bethesda MD United States Public Health Service 1999

119 National Heart Lung and Blood Institute Strategic Plan FY 2002ndash06 Rockville MD National Institutes of Health 2002 httpwwwnhlbinihgovresourcesdocsplanindexhtm accessed onApril 1 2003

120 Mangione C Reynolds E Disparities in health and health care JGen Intern Med 200116276ndash80

121 Weisse C Sorum P Sanders K Syat B Do gender and race affectdecisions about pain management J Gen Intern Med200116211ndash7

122 Chapman K Tashkin D Pye D Gender bias in the diagnosis ofCOPD Chest 20011191691ndash5

123 Watson R Stein A Dwamena F et al Do race and gender influencethe use of invasive procedures J Gen Intern Med 200116227ndash34

124 Raine R Does gender bias exist in the use of specialist health careJ Health Serv Res Policy 20005237ndash49

125 Johnson M Lin M Mangalik S Murphy D Kramer A Patientsrsquoperceptions of physiciansrsquo recommendations for comfort care differby patient age and gender J Gen Intern Med 200015248ndash55

126 Majeed Z Moser K Maxwell R Age sex and practice variationsin the use of statins in general practice in England and Wales JPublic Health Med 200022275ndash9

127 Savoie I Kazanjian A Utilization of lipid-lowering drugs in menand women a reflection of the research evidence J Clin Epidemiol20025595ndash101

128 Hippisley-Cox J Pringle M Crown N Meal A Wynn A Sexinequalities in ischaemic heart disease in general practice cross-sectional survey BMJ 2001322832

129 Bowker T Clayton T Ingham J et al British Cardiac Societysurvey of the potential for the secondary prevention of coronarydisease ASPIRE (Action on Secondary Prevention through Inter-vention to Reduce Events) Heart 199675334ndash42

130 Bannerman A Hamilton K Isles C et al Myocardial infarction inmen and women under 65 years of age no evidence of gender biasScott Med J 20014673ndash8

131 Wei L Wang J Thompson P Wong S Struthers A MacDonald TAdherence to statin treatment and readmission of patients aftermyocardial infarction a six year follow-up study Heart200288229ndash33

132 Sgadari A Incalzi R Onder G Pedone C Gambassi G Lipid-lowering therapy in patients with coronary artery disease sex orage bias Arch Intern Med 20001602684ndash5

133 Pilote L Beck C Richard H Eisenberg M The effects of cost-sharingon essential drug prescriptions utilization of medical care andoutcomes after acute myocardial infarction in elderly patientsCMAJ 2002167246ndash52

134 Evenson K Fleury J Barriers to outpatient cardiac rehabili-tation participation and adherence J Cardiopulm Rehabil200020241ndash6

JGIM


855

procedures Although the literature documenting suchdisparities is extensive again the mechanisms remainvirtually unexplained and efforts to explain them havetended to focus on clinician factors

16ndash23

Elucidating themultiple factors that contribute to gender disparities mayhelp structure quality improvement interventions for bothmen and women

MODEL

Patient clinician and health system factors may allcontribute to suboptimal management of dyslipidemiaDrawing from the Health Belief model

24

the Landon et alhealth care organization model

25

and the Jaen et al com-peting demands model

26

we have developed a con-ceptual model (Fig 1) for how these factors interact Atheoretical model should have face validity provide mea-surable variables and enhance understanding beyondwhat would be expected from consideration of individualfactors affecting preventive service delivery In our concep-tual model we have incorporated the concept of patientperception of risk and subsequent behavior from the HealthBelief model the association between detailed structuralcharacteristics of health systems and physician behaviorfrom Landonrsquos model and the role of physician character-istics and the idea of competing interests from the Jaenet al model

The Health Belief model argues that health behaviorsare related to personal beliefs about susceptibility to dis-ease seriousness of disease benefit of intervention andrisk of intervention In this model individuals who do notbelieve they are at high risk of disease are unlikely to pur-sue preventive health behavior even if the benefits of thebehavior are high and the risks are low and individualswho believe they are at high risk of disease may pursuepreventive health behavior even if the benefits are low andthe risks are high

24

This model has proven useful in under-standing and predicting many preventive health behaviorsincluding diet and exercise However the model is not asuseful in examining the barriers to acting on such beliefsie barriers related to the structural organization of thehealth care system and barriers related to specific aspectsof the health care visit such as limited time In the HealthBelief model the barriers are limited primarily to thepatientrsquos perceived barriers to behavior change

In contrast the Landon et al model of health care organ-ization focuses on characteristics of the health care systemthat can influence health care delivery

25

In the Landon et almodel disease processes and outcomes can be influencedby financial incentives management strategies such as utili-zation review structure of care such as the location of thepractice site and staffing patterns and finally normativeinfluences such as the culture of the organization Thestrengths of this model are that it details health plan and

Table 1 Studies That Reported Results by Gender in Screening or Treatment for Dyslipidemia in Patients with Cardiovascular Disease

Data Source Results Extent of Disparity

Randomized trial56 Women less likely to receive lipid treatment and to be at goal than men

35 vs 55 (P lt 05) 29 vs 48 (P = 001)

Registry59 Women less likely to have lipid levels measured than men

Not given

Registry49 Women less likely to receive lipid treatment and to be at goal than men

44 vs 45 95 vs 161 (P value not given)

Registry1213 Women less likely to be at treatment goal than men

P lt 001 (percents not given)

Registry126 Women less likely to receive statins than men 82 vs 133 (P value not given)Pharmacy database127 Women less likely to receive statins than men 135 vs 208 (P lt 05)Patient survey60 Women less likely to receive lipid treatment and

to be at goal than men 298 vs 396 203 vs 219

(P value not given)Electronic medical record54 Women less likely to be screenedtreated

than men243 vs 373 (P lt 0001)

Chart review48 Women less likely to be treated than men 33 vs 48 (P = 047)Chart review95 Women less likely to be at lipid goal than men P lt 05 (percents not given)Chart review128 Women less likely to be screened treated

and at goal than men 35 vs 50 21 vs 31 23 vs 33

(P lt 0001)Chart review97 Women less likely to be screened and

at lipid goal 51 vs 68 (P = 001) 25 vs 34

(P = 043)Randomized trial58 No gender difference in screening or treatmentPatient survey129 No gender differences in screening or treatmentChart review57 No gender differences in treatmentChart review130 No gender difference in screening or treatmentPharmacy database131 No gender difference in treatmentPharmacy database132 No gender difference in treatmentPharmacy database133 No gender difference in treatmentPatient survey80 Women more likely to be treated 64 vs 81 (P lt 05)

856


JGIM



26






Access to Care


27


28



JGIM


857


29


3031


Cost Barriers


32


33


27


34



1435



3637


38


3940


26


4142


4344

Demographics



References





858



Disease Severity






Other Confounders












Prioritization
















CONCLUSIONS



















































































































































856


JGIM



26






Access to Care


27


28



JGIM


857


29


3031


Cost Barriers


32


33


27


34



1435



3637


38


3940


26


4142


4344

Demographics



References





858



Disease Severity






Other Confounders












Prioritization
















CONCLUSIONS



















































































































































JGIM


857


29


3031


Cost Barriers


32


33


27


34



1435



3637


38


3940


26


4142


4344

Demographics



References





858



Disease Severity






Other Confounders












Prioritization
















CONCLUSIONS



















































































































































858



Disease Severity






Other Confounders












Prioritization
















CONCLUSIONS























































































































































Prioritization
















CONCLUSIONS
























































































































































CONCLUSIONS



















































































































































Review of evidence and explanations for suboptimal screening and treatment of dyslipidemia in women

Documents

Transcript of Review of evidence and explanations for suboptimal screening and treatment of dyslipidemia in women