Prevalence of True Therapeutic Inertia in Blood Pressure Control in an Academic Chronic Kidney Disease Clinic


Address for correspondence: Nayan Desai, MD, Department of Medicine, Cooper University Hospital, E & R Building, 401 Haddon Avenue, Third floor, Camden, NJ 08103



Therapeutic inertia (TI) in blood pressure (BP) control has been traditionally defined as failure to initiate or intensify therapy when treatment goals are not met. The fallacy with this definition is that TI may be overestimated because it includes hypertensive patients deliberately uncontrolled. This is a retrospective chart review study that evaluated physicians' response to an uncontrolled clinic BP reading in a population of patients with stage 3 to 5 chronic kidney disease (CKD) and hypertension. Of 429 patients screened, 166 had controlled BP and 263 did not. Of these 263 patients, 115 patients had no clear reason documented for the absence of changes in medication regimen. This population was defined as cases with true TI. In the remaining 148 patients, the medication regimen was changed in 81 patients. In the rest of the patients, there was a reason documented for not changing the medication regimen. The prevalence of true TI rate (defined as percentage of uncontrolled hypertension as a result of physician inaccountability) in our study was 44% as compared with 69% if the traditional TI definition is applied. Thus, we conclude that the prevalence of TI in the literature overestimates the rate of true TI as it does not account for physician decision making. The current definition of TI in BP control needs to be revised, as it underestimates a provider's care to improve BP control and is misleading. The TI definition should include some mechanism to account for interventions beyond medication titration.

Hypertension (HTN) is a chronic problem that affects more than 67.5 million adults in the United States.[1] As per the recent National Health and Nutrition Examination Survey (NHANES), antihypertensive medication use among US adults with HTN has increased from 63.5% to 77.3% during the past decade (2001–2010). However, there continues to be a gap between HTN treatment and control, as blood pressure (BP) was controlled in only 47% of all hypertensive patients and 60% of treated hypertensive patients.[2] The reasons for poor BP control are multifactorial and include patient, physician, and care delivery factors.[3] Of these, suboptimal therapy has been identified as an important physician factor. This is known as clinical inertia or, as described by Phillips and colleagues,[4] as “failure of health care providers to initiate or intensify therapy when indicated.” Clinical inertia has also been called therapeutic inertia (TI). Despite the increase in the number of hypertensive medications available and well-published guidelines regarding BP control targets, clinical inertia is present in two thirds of hypertensive visits.[5] This occurs despite patients being seen, when BP is uncontrolled, on average 6 times annually by their primary care providers.[6]

Studies on TI have focused on failure to introduce or intensify antihypertensive therapy by providers in uncontrolled hypertensive patients. This definition, however, excludes nonpharmacologic interventions and deliberately uncontrolled hypertensives. Physicians may recommend lifestyle modifications such as sodium restriction or weight loss. On similar grounds, the provider may have a valid reason for not intensifying therapy based on the clinical setting. These include doubts about the validity of the office BP obtained (due to home readings being different or missed medications) as well as concerns regarding the risks of medication titration, including orthostasis. Low diastolic BP (DBP) is a common scenario that limits titration of medications when systolic BP (SBP) is not at goal.

Patients with chronic kidney disease (CKD) with HTN are a specific subset of patients who may benefit from tight BP control, but who also may have detrimental effects from tight BP control. Examples of these include worsening renal function, orthostasis, and low DBP due to a large pulse pressure. Prior studies have not looked into the prevalence of true TI, which excludes patients with nonpharmacologic intervention and deliberately uncontrolled HTN. Taking into consideration these factors, we conducted a retrospective chart review study in hypertensive CKD patients in an academic CKD clinic to identify the prevalence of true TI.


Principle Hypothesis

We evaluated nephrologists' response to uncontrolled BP in a hypertensive CKD population in an academic clinic. Our primary hypothesis was that TI in BP control is present in 40% of patients in an academic CKD clinic but a significant proportion of these are “deliberately” uncontrolled and thus do not truly represent TI.

Study Setting and Population

This is a retrospective cross-sectional study of patients with HTN and CKD at Cooper University Hospital CKD outpatient clinics. We identified patients from electronic medical records based on International Classification of Diseases code for CKD stage 3, 4, and 5 (585.3-5). Electronic medical records of identified patients from October 2009 to December 2011 were reviewed. All patients were older than 18, had an estimated epidermal growth factor receptor (eGFR) <60 mL/min as estimated by the 4-variable Modification of Diet in Renal Disease equation, and had stage 3 to 5 CKD. Patients with end-stage renal disease on dialysis were excluded from the study. Details of the most recent visit (one visit per patient) seen by a nephrologist were recorded. Seven nephrologists at our academic institute are involved in the management of hypertensive CKD patients. Therefore, we also recorded the name of the nephrologist for each visit to estimate the prevalence of TI for each nephrologist.

The following characteristics for each patient were recorded in a database: age, sex, presence of diabetes, total number of antihypertensive medications and type of BP medications, pulse pressure, systolic and diastolic BP, creatinine, and CKD stage as described in the office note. For patients with more than one BP reading, the second reading was entered in the database. Based on the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC 7) guidelines for BP control, patients with BP <130/80 mm Hg were labeled as controlled.[7] Patients with BP ≥130/80 mm Hg were labeled as uncontrolled. In the uncontrolled hypertensive patients, one of the following actions was assigned based on a review of the clinic visit note:

  • Titration of medications
  • No new medications given as concerns of clinic BP being inaccurate (includes medications missed or use of home BP)
  • No new medication given because of orthostatic symptoms or history of falls
  • No new medication given because of low DBP
  • No new medication given; dietary and lifestyle modifications given for BP control
  • No new medication given because of patient refusal
  • No new medication given; no comment on HTN not at goal.

We defined Category vii as true TI for the purpose of this study. If the physician had not commented on patients' uncontrolled BP for that particular visit, it was recorded as no comments or true TI. Institutional review board approval was obtained for the study.

Sample Size Rationale and Statistical Methods

We determined that 263 uncontrolled patients would provide 90% power to detect a deliberately uncontrolled prevalence rate of 45% to 55%. The prevalence rate is based on data from studies in hypertensive CKD patients.[8] Student t test was used to compare numerical continuous variables including age, BP, creatinine, and CKD stage. Chi-square test was used to compare categorical variables including sex, diabetes, and antihypertensive therapy.


A total of 508 charts were reviewed, of which 429 patients satisfied the inclusion criteria. Seventy-nine patients who were excluded were either inpatients (50), did not have the diagnosis of HTN (12), or had end-stage renal disease (7).

As determined, 263 patients had uncontrolled BP (≥130/80 mm Hg), and the remaining 166 patients had controlled BP. The mean age of the studied population was 69.9 years, and 48% of the patients were men. Diabetes as a comorbidity was present in 48% of uncontrolled patients. Of 81 patients with stage 4 CKD, 11 patients were taking thiazide diuretics and of 19 patients with stage 5 CKD, only 2 patient was taking a thiazide. In comparison, 68 of 238 patients with stage 3 CKD were taking thiazide diuretics. Patients with uncontrolled BP were taking 2.7 antihypertensive agents as compared with 2.4 agents in the controlled group. In the uncontrolled group, 63.4% of patients were taking a β-blocker, 45.5% were taking an angiotensin-converting enzyme inhibitor or angiotensin receptor blocker, and 44.8% were taking a calcium channel blocker. The mean systolic and diastolic BP in the controlled group were 115.2 mm Hg and 64.7 mm Hg, respectively, whereas in the uncontrolled group they were 141.6 mm Hg and 75.6 mm Hg, respectively. Race was not recorded in the majority of the patient charts and, as a result, eGFR could not be recorded. Table 1 highlights the study characteristics of both the uncontrolled and controlled groups. There was a statistically significant difference seen in the mean number of total antihypertensive agents and the mean creatinine and CKD stage among both the groups.

Table 1. Characteristics of Patients in the Controlled and Uncontrolled Groups
CharacteristicsControlled Group (n=166)Uncontrolled Group (n=263)P Value
  1. Values are presented as number (percentage) unless otherwise indicated. Abbreviations: ACE, angiotensin-converting enzyme; ARBs, angiotensin receptor blockers; CKD, chronic kidney disease; Cr, creatinine; DBP, diastolic blood pressure; SBP, systolic blood pressure; SD, standard deviation.

Age, mean (SD), y69.92 (12.4)69.81 (12.5).93
Men 94 (57)125 (48).07
Diabetes mellitus72 (43)126 (48).34
Calcium channel blockers57 (49)118 (45).02
β-Blockers100 (86)167 (63).48
Thiazides34 (29)63 (24).40
Loop diuretics70 (60)117 (44).54
Aldosterone antagonists16 (14)23 (9).76
ACE inhibitors40 (34)62 (24).91
ARBs35 (30)58 (22).77
Total agents, mean (SD)2.4 (1.4)2.7 (1.3).05
Pulse rate, per min, mean (SD)72.86 (12.2)75.34 (12.6).05
SBP, mean (SD), mm Hg115.2 (9.7)141.61 (14.3) 
DBP, mean (SD), mm Hg64.7 (8.2)75.6 (10.8) 
Cr in mg/dL, mean (SD)1.98 (1)2.2 (1.1).04
CKD stage, mean (SD)3.2 (0.5)3.3 (0.6).01

In the 263 patients with uncontrolled BP, 115 (44%) patients had no comments in the electronic medical charts for their uncontrolled BP; hence, this was determined to be true TI. The prevalence of true TI ranged from 21% to 59% among the 7 nephrologists in the study group. In the remaining 148 patients, medications were intensified in 81 patients. In the remaining 67 patients (deliberately uncontrolled group), there was a reason documented in the chart for not introducing or intensifying drug therapy. Comparing the mean age, total number of antihypertensive agents, creatinine, and CKD stage, both groups (deliberately uncontrolled and true TI) were similar and the P value was not statistically significant (Table 2). Patients in the deliberately uncontrolled group had higher mean systolic and diastolic BPs compared with patients in the true TI group (P<.01).

Table 2. Characteristics of the Uncontrolled Group: “Deliberately Uncontrolled” Group and “True Therapeutic Inertia” Group
CharacteristicsDeliberately Uncontrolled Group (n=148)True Therapeutic Inertia Group (n=115)P Value
  1. Values are presented as number (percentage) unless otherwise indicated. Abbreviations: ACE, angiotensin-converting enzyme; ARBs, angiotensin receptor blockers; CKD, chronic kidney disease; Cr, creatinine; DBP, diastolic blood pressure; SBP, systolic blood pressure; SD, standard deviation.

Age, mean (SD), y69 (13.1)70.8 (11.8).25
Men79 (53)46 (40).03
Diabetes mellitus 79 (53)47 (41).05
Calcium channel blockers 68 (46)50 (43).70
β-Blockers94 (63)73 (63).94
Thiazides 28 (18)35 (30).02
Loop diuretics 66 (46)51 (44).93
Aldosterone antagonist 14 (9)9 (8).64
ACE inhibitors42 (28)20 (17).05
ARBs34 (23)24 (21).69
Total agents, mean (SD)2.8 (1.8)2.6 (1.1).27
Pulse rate, per min, mean (SD)75.1 (11.8)75.5 (13.6).81
SBP, mean (SD), mm Hg146.6 (14.5)135.1 (11.1)<.01
DBP, mean (SD), mm Hg77.6 (10.4)73.0 (10.7)<.01
Cr, mean (SD), mg/dL2.2 (1.1)2.2 (1).97
CKD stage, mean (SD)3.3 (0.6)3.3 (0.6).80

Of the 148 patients without true TI, new medications were introduced or existing medications were titrated in 81 patients (Figure 1). In 67 patients, lifestyle modifications (salt-restricted diet, exercise) were discussed with the patients. Medical therapy was not intensified in 20 patients, as office BP was felt to not be representative of patient's true BP. The reasons for this include white-coat HTN, home BP or ambulatory readings showing well-controlled BP, or if patients skipped their medications on the day of office visit. Similarly, medical therapy was not prescribed in 9 patients because they refused new medications. Orthostatic BP changes or symptoms of orthostasis were present in 4 patients. Two patients did not warrant treatment secondary to their low DBP. This number exceeds the actual number of patients since more than one intervention may have been performed.

Figure 1.

Uncontrolled blood pressure (BP). DBP indicates diastolic BP.

We further studied the systolic and diastolic BP characteristics of patients with uncontrolled BP. As seen in Figure 2, nephrologists commented on 95 patients (64%) with uncontrolled SBP if it was >140 mm of Hg but less frequently when SBP was <140 mm Hg. DBP did not follow the similar trend as the majority of the patients in both groups had controlled DBP.

Figure 2.

Systolic blood pressure (SBP) among the “deliberately uncontrolled” group and “true therapeutic inertia” group.


In this observational study, we studied 429 patients with CKD and HTN under the care of a nephrologist in an academic nephrology clinic. We studied the prevalence of TI among 263 uncontrolled hypertensive patients. We attempted to define reasons for lack of titration of antihypertensive therapy. Specifically, we aimed to determine whether the frequent lack of interventions for uncontrolled BP represented active recognition regarding the lack of BP control, previously described as deliberately uncontrolled, or represented a failure to respond to lack of BP control, described as “true therapeutic inertia.” We found that the prevalence of true TI in our study population was 44%. If the conventional definition of TI is applied, this number increases to 69% (Figure 3). Hence, BP medication titration is not the only action taken by physicians for uncontrolled BP. The main finding is that interventions to control BP are not only limited to medication titration.

Figure 3.

Pie chart of uncontrolled hypertensive patients. TI indicates therapeutic inertia.

Historically, BP control is far from ideal in the CKD population. Specifically, achievement of BP targets of <130/80 mm Hg was achieved in 37%, 13%, and 46% of CKD patients from 3 studies, namely NHANES, Kidney Early Evaluation Program (KEEP), and Chronic Renal Insufficiency Cohort (CRIC) studies, respectively.[8-11] In our study, this rate was 39% (166 of 429 patients). The mean age of patients in the uncontrolled group in our study was 69.8 years, which was higher than that in NHANES (2005–2008); the sex distribution was similar. The mean systolic and diastolic BP in our study was 141.6 mm Hg and 75.6 mm Hg, which was comparable with 150.5 mm Hg and 79.7 mm Hg in the NHANES 2005–2008 cohort.[8] TI in NHANES was present in more than two thirds of the uncontrolled hypertensive population, which is similar to our rate of 69% before accounting for the deliberately uncontrolled patients.

Although thiazide and thiazide-like diuretics are deemed ineffective in stage 4 CKD, their use has been debated in the recent literature. In our study population, 13.6% and 5.3% patients with CKD stage 4 and 5, respectively, were taking thiazide diuretics. Review of the literature suggests that thiazides may be useful among patients with stage 4 CKD and above; however, randomized control trials are needed before their use is recommended.[12]

Nephrologist were more likely to intervene if SBP was >140 mm Hg or if patients had a higher CKD stage. The majority of patients with true TI had their BP within 10/5 mm Hg of goal BP. It is very likely that being so close to goal was the reason for lack of intervention. When the group whose BP was <140/80 mm Hg was not included, the true TI rate dropped to 23% (Figure 2). We believe this represents physician determination that an individual's cardiovascular risk profile may not be improved with a BP <140/80 mm Hg in these select cases. Previous nephrology literature supports concern for the lack of clinical data justifying lower BP in all patients.[13]

A growing body of literature supports the concept that out-of-office BP is a better predictor of future cardiovascular events than office BP in the CKD as well as non-CKD populations.[14] Thus, it is reasonable for BP therapy to be guided by non–office BP presuming there is some suggestion that BP is better controlled than documented in the clinic.[15] In our study, this clinical uncertainty was present in 7% of uncontrolled hypertensives; hence, therapy was not intensified. It would be incorrect to classify these patients as experiencing TI. Home BP or ambulatory BP measurement may provide an answer in case of this uncertainty.[16]

A single BP reading documented in the chart was used to define control vs uncontrolled BP. No repeat BPs were reported, which is not consistent with guideline recommendations. As per the American Heart Association and JNC 7 guidelines at least two BP readings should be made and the average reading recorded. As observed in our study group, having only a single reading reported in the chart at an academic center underscores the need for renewed education regarding BP measurement.[7, 17]

A reduction in dietary sodium intake leads to reduced BP, thus therapeutic lifestyle measures form an important backbone in the treatment of HTN.[18] This would include diet, exercise, and weight loss in addition to a reduction in sodium intake. These recommendations were provided to this population to reduce BP but would have been unappreciated if traditional TI definitions were used. Failure to emphasize and prescribe these measures is defined as behavioral inertia.[19] Lifestyle modifications in our study were prescribed in 25% of patients.

There are limited high-quality data to suggest a BP goal <130/80 mm Hg is beneficial. In addition, concern exists for the possibility that an excessively low diastolic BP may increase cardiovascular events.[20]

Similarly, orthostatic hypotension in CKD patients may prevent intensification of therapy. In our study, medical therapy was not intensified in 2 and 5 patients for concern of diastolic hypotension and orthostatic hypotension, respectively. What could have been labeled as TI, is truly a clinical safeguard for patients with life-threatening complications.[21]

Lastly the issue of medication noncompliance contributing to uncontrolled BP is also included in the current definition of TI. Intensifying medical therapy has no role in patients who are nonadherent to therapy. Root cause analysis of the problem for nonadherence may help to achieve better BP control.

Although the rate of true TI is one fourth less than conventional TI, it is still high at 44%. BP control is still inadequate, and substantial TI does exist in this subset of population studied.

Study Strength and Limitations

This study enrolled patients seen in a specialty nephrology clinic at an academic center. BP was frequently not controlled, consistent with other large-scale studies. The nephrologists' subsequent response, in order to provide the best patient care, provides the data from which the conclusions are drawn. In that sense, the study shows a potential downside of current database studies of TI. An important fact highlighted from our study is the way database studies are conducted in the current era. With the growing number of electronic databases, conclusions can be drawn with a click of a mouse. Individual patient charts are not reviewed once the databases are created as they are time-consuming; this may result in missing a key nontherapeutic intervention that might have been documented in patient records. This brings into question the reliability of the results of the previous database studies in TI.

We believe that reviewing individual patient charts, though a time-consuming process, gave edge to our study in estimating the prevalence of both true and conventional TI over other database studies. Additionally, as a retrospective study, the physicians were not aware that their medical decision making would be critiqued and, as a result, such bias was removed.

Our investigation has the limitation of being a single-center observation study and therefore its applicability is certainly not guaranteed. In addition, the results were based on a single visit and one BP reading in the clinic. An additional limitation is that outcomes were based on documentation in patient charts. Interventions undertaken by physicians with regards to BP control could be underestimated if not documented. Also, the findings are relevant to similar HTN patients with CKD and should not be extrapolated to hypertensive patients without kidney disease or those undergoing dialysis.


We conclude that the current definition of TI needs to be revised. Specifically, some method needs to be present for interventions that do not lead to BP medication titration but that do either lead to better BP control, as in the example of sodium intake restriction, or provide potentially better care, as in avoiding medication titration in orthostatic patients. Lifestyle modifications, clinical uncertainty of office BP, medication noncompliance, and diastolic and orthostatic hypotension are situations that may currently be associated with TI. We believe this underestimates provider care to improve BP control and is misleading. Thus, the term “therapeutic inertia” should be revised to include some mechanism to account for interventions beyond medication titration.

Acknowledgments and disclosures

The authors report no specific funding in relation to this research and no conflicts of interest to disclose.