As the US population ages, increasing numbers of individuals have multiple medical conditions [1, 2]. As a result, primary care clinicians are increasingly seeing patients with multiple conditions, such as diabetes and/or hyperlipidemia, that contribute to increased risk of cardiovascular disease (CVD). Although the two conditions frequently occur in the same patient, little information is available regarding how to prioritize disease management and implement guideline-concordant care for complex patients with multiple uncontrolled conditions [1–3]. In practice, clinicians manage uncontrolled conditions inconsistently at every clinical encounter [4–7], and rates of simultaneous control of all co-existing conditions are suboptimal [8].

The phenomenon of patients presenting for care with uncontrolled conditions or risk factors and the clinician not adjusting the medication regimen has been called 'clinical inertia' [9, 10]. In primary care practice, there are situations in which clinicians may choose not to follow disease-specific guidelines or adhere to quality measures based on clinically appropriate reasons. While reports of clinical inertia are accumulating, the reasons why it occurs remain limited [11]. In addition, it is unknown whether the presence of multiple uncontrolled conditions contributes to clinical inertia.

To examine the role of multiple uncontrolled conditions in clinical inertia, we studied the association of blood pressure (BP) intensification and the presence and control of two other CVD risk factors, diabetes and hyperlipidemia, in patients with uncontrolled hypertension. We hypothesized that additional uncontrolled CVD risk factors would act as competing demands, decreasing the likelihood that clinicians would adjust BP medications when the patient's BP was above goal.

In this study, patients with multiple uncontrolled CVD risk factors were more likely to have BP medication intensified, contrary to our expectations. Our hypothesis that multiple uncontrolled CVD risk factors, specifically diabetes and hyperlipidemia, would act as competing demands and thereby distract clinicians from treating elevated BP was not supported. Higher BP medication intensification rates in patients with multiple uncontrolled conditions suggest that clinicians may be appropriately identifying and managing high-risk patients with uncontrolled hypertension, those patients with higher CVD risk who are more likely to benefit from BP control.

Our results are concordant with other studies indicating better quality of care for medically complex patients [13, 14]. Heisler et al. also reported higher medication intensification rates among patients with higher BPs [15]. We found that other uncontrolled CVD risk factors had a similar effect and did not in fact 'compete' for the clinician's attention to BP management. Because many of these patients had additional medical conditions besides diabetes and/or hyperlipidemia, further investigation is needed to assess the effect of conditions unrelated to CVD risk on decision-making and disease management [16]. Particularly, it is not clear how painful arthritis, for example, influences CVD risk factor management.

These findings reinforce our and others' work suggesting that inaction by clinicians is not necessarily clinical inertia [11, 17]. Inaction may include both appropriate and inappropriate clinical care. Appropriate inaction, such as not intensifying medications in the setting of non-adherence, reflects the challenge of applying guidelines to the care of individual patients balanced with meeting population-based quality indicators. We noted a considerable amount of appropriate action in the situation of medication intensification for the highest risk patients (BP at least 10 mmHg above goal and uncontrolled diabetes and hypertension).

On the other hand, intensification rates observed when the BP was closer to goal were quite low and may represent appropriate inaction. In these modest ranges of elevation, clinicians may be more cautious, possibly weighing the modest benefits against the potential harms of a more intense regimen. The difference in management of patients close to BP goals and those with higher BP highlights the intricacy of providing guideline-concordant yet patient-centered care in primary care. Additionally, BP measurement is known to be imprecise, and low elevations may require close monitoring to assess the stability of the level. Further, the risks of polypharmacy are real and an important potentially appropriate reason to forego intensification of the BP regimen. In addition, a single BP level does not reflect past efforts, which may have resulted in considerable BP lowering and cardiovascular risk reduction. Unfortunately, dichotomous quality indicators, including the one assessing BP control, do not acknowledge the subtleties involved in 'real world' clinical decision making, nor do they account for appropriate inaction. In fact, tying rewards to a simplistic quality indicator that overlooks potentially clinically appropriate inaction may result in unintended consequences, creating incentives for polypharmacy and raising costs.

We observed that clinicians were possibly inappropriately overestimating the benefit derived from controlled diabetes and/or hyperlipidemia. BP medication intensification rates were consistently the lowest for individuals with either controlled diabetes or hyperlipidemia even with systolic BP > 10 mm Hg above goal. These rates were lower than for individuals with only uncontrolled hypertension, and including the number of hypertension, diabetes, and hyperlipidemia medications in the model did not change this effect. Controlled diabetes and/or hyperlipidemia lowers but does not eliminate the risk of CVD, so efforts to improve BP control remain important in reducing CVD risk [18]. In light of recent data from the ACCORD study, control of systolic BP to 120 mm Hg failed to demonstrate a decrease in major cardiovascular events in patients with diabetes compared to less stringent BP (systolic BP < 140 mm Hg) [19, 20]. Regardless, lack of treatment intensification when BP is notably above goal among patients with known but controlled diabetes and/or hyperlipidemia could be the target of quality improvement activities, and may represent true clinical inertia.

Our study of clinicians who practice in a healthcare system with strong quality improvement initiatives and incentives to control BP allowed us to observe whether clinicians implement guidelines differently based on individual patient factors, including the benefits and risks of medication intensification. These individual patient factors and complex clinical decisions may not be apparent by measuring quality indicators using administrative data and may only be documented with observational data.

Our study has several limitations. Although the sample of 13 clinicians at one VA facility caring for mostly male patients raises issues of generalizability, our results confirm past reports of better quality of care with more medically complex patients. With only 13 participating clinicians, we were not able to explore clinician characteristics that could be contributing to variations in care, and this may warrant further investigation with a larger study. Additionally, we may have seen higher rates of medication intensification than typical practice settings because clinicians knew we were asking them to document whether they changed medications. Our sample was too small to confirm our preliminary findings suggesting that BP medication intensification patterns were similar for controlled versus uncontrolled hyperlipidemia and for controlled versus uncontrolled diabetes. We were also unable to conduct multivariate analyses of these findings, which differ from past reports indicating that hyperlipidemia is more often ignored compared to uncontrolled diabetes or hypertension [21, 22]. These findings require confirmation in larger studies. Important patient-level factors that could contribute to suboptimal BP control and clinical inaction, such as cost or adherence issues, were not available.

The presence of uncontrolled diabetes and hyperlipidemia increased the likelihood of BP medication intensification in patients with uncontrolled hypertension. This effect was observed regardless of the level of BP elevation, but the magnitude of the effect was greater if BP was further from goal. These findings suggest that clinicians correctly identified and treated the highest risk hypertensive patients likely to receive the greatest benefit with BP medication intensification. Lower BP medication intensification rates seen in patients with modest BP elevations and in patients with controlled diabetes and/or hyperlipidemia may represent appropriate clinical inaction and inappropriate clinical inertia respectively, but further investigation is needed. Our study reflects the complex decision making process involved in providing medical care that balances clinical guideline-concordance (based on the benefit to groups or populations) and patient-centeredness to patients with multiple uncontrolled conditions. The implementation of quality indicators that measure medication intensification, account for the clinical risks and benefits of medication intensification, and assess both appropriate clinical actions and inactions are needed to maximize high quality, patient-centered care.