#NephJC Chat
Tuesday, Sept, 5th, 2023 at 9 pm Eastern (AEST = Sept 6th, 11 am)
Wednesday, Sept 6th, 2023, at 9 pm Indian Standard Time and 3:30 pm GMT (AEST = Sept 7th, 1:30 am)
JAMA Intern Med 2023 Jul 1; 183(7):715-723.
doi: 10.1001/jamainternmed.2023.1667
Clinical Outcomes of Intensive Inpatient Blood Pressure Management in Hospitalized Older Adults
Timothy S Anderson, Shoshana J Herzig, Bocheng Jing, W John Boscardin, Kathy Fung, Edward R Marcantonio, Michael A Steinman
PMID: 37252732
Introduction
Most of us, at some point while covering an inpatient service, have been called for an asymptomatic elevated blood pressure (BP) reading and reflexively responded with an order to administer antihypertensives. Perhaps you felt pressured to “normalize” the patient’s pressure, but is that really the right strategy? Some might argue blood pressure is measured in asymptomatic hospitalized patients primarily for the purpose of detecting hypotension, and that hypertension in the hospitalized patient is inconsequential. The prevalence of inpatient hypertension is as high as 50 to 70% (Axon RN et al, J Hosp Med 2011) in some studied populations. In addition, 28 to 38% of patients have elevated blood pressure readings in the absence of a diagnosis of hypertension prior to hospitalization (Conen D et al, J Hypertens 2006). The high prevalence of blood pressure elevation in inpatients could be situational—related to pain, anxiety, sleep disruption, dietary and mobility restrictions, and the stress of an unfamiliar environment.
Image created by AI, based on prompts by Jithu Kurian
Although we have a standardized way of measuring blood pressure in the outpatient setting and numerous thoughtful medical society guidelines, no such procedures or goals exist for patients in the hospital setting (Axon RN, et al, J Clin Hypertens (Greenwich), 2010). Obviously, hospitalized patients with elevated blood pressure comprise a diverse group of underlying comorbidities and etiologies for hypertension, with a plethora of options for therapeutics as well. As virtuoso nephrologists acquainted with hypertensive management, one might wonder if there is a specific population or therapeutic strategy that is helpful to hypertensive hospitalized patients.
To date, multiple observational studies have found no beneficial effects of intensive blood pressure lowering in patients hospitalized with non-cardiovascular (or non-pregnancy) related diagnoses.
In fact, intense inpatient blood pressure treatment has been observed to be associated with an increased incidence of acute kidney injury (AKI) and myocardial injury (MI) (Rastogi R et al, JAMA Intern Med 2021). In another study, researchers compared “as needed” intravenous labetalol or hydralazine administration for inpatients whose blood pressures were above a specific threshold. Unfortunately, this strategy was found to actually increase the inpatient length of hospital stay (Bean-Thompson K et al, Am J Hypertens 2021). Finally, another observational study found that patients who received any hypertensive medications ‘PRN’ while hospitalized had a higher incidence of mortality, AKI, and stroke compared to patients who received only scheduled antihypertensives (Mohandas R et al, Hypertension 2021). So far, in most studied inpatient populations, there has been no compelling evidence that treating elevated blood pressure in the hospital improves outcomes.
This latest study, by Anderson et al, looked at the composite outcomes of inpatient mortality, ICU transfer, stroke, AKI, BNP elevation, and troponin elevation in elderly inpatients whose blood pressures were “intensively controlled” (as defined below) as compared to those who received “non-intense” blood pressure management. The study was innovative in looking at intensive blood pressure management in elderly patients, who are most susceptible to adverse events and adopted a target trial emulation design, which differs from previous observational studies.
The Study
Methods
Study design
This study was a retrospective observational cohort study conducted from October 1 2015 to December 31 2017 within the United States Veteran Affairs system. Data was collected using inpatient and outpatient clinical and pharmacy information from the Veterans’ Health Administration (VHA) via claims from 2013 to 2018.
The study aimed to examine the association between intensive inpatient treatment of elevated BPs and in-hospital clinical outcomes among older adults admitted for non-cardiovascular conditions.
Patient selection
The study included patients:
Aged ≥65 years hospitalized at a VHA hospital with a non-cardiovascular diagnosis AND
Who had two or more elevated blood pressure readings (defined as an SBP ≥140 mmHg) within the first 48 hours of hospitalization (timed from the first measurement of vital signs in the hospital)
The following patients were excluded from the study:
Patients admitted for cardiovascular disease, cerebrovascular conditions, acute renal failure, symptoms related to hypertensive emergencies
Patients who were discharged or experienced a study outcome within 48 hours of admission
Patients admitted to intensive care, surgical, psychiatric, inpatient rehabilitation, or another acute care hospital
Patients with ESKD prior to hospitalization
Patients admitted from a nursing home who availed less than 80% of their outpatient visits in the VHA system and those who received 0 drugs from the VHA pharmacy in the year prior to hospitalization
The observational study was designed to emulate a target trial of intensive versus non-intensive inpatient antihypertensive treatment in the elderly. How does this help in clinical decision making?
Even randomized control trials (RCT) have potential biases (i.e. immortal, lead time, and depletion of susceptibility bias). These errors are completely preventable by using target trial emulation, which reproduces the randomized control trial design. Three components are necessary at randomization: eligibility criteria met, assignment to a treatment strategy, and start of follow-up (time zero). These three elements actually prevent immortality bias and confounding bias. Thus, observational studies can complement RCTs to inform clinical practice, but only if they are done properly to prevent additional errors. Thus, an emulation of a target trial is the application of principles from RTCs to the analysis of observational data, thereby tying the analysis to the trial it is emulating (for more details on target emulated trials see the #NephJC about dialysis timing in advanced CKD, and this JASN review explaining type of errors target emulated trials bust).
Interventions:
Intensive BP treatment was defined as use of:
IV anti-hypertensive or
A new oral medication not previously prescribed prior to admission
Outcomes
The primary outcome was a composite of inpatient mortality, AKI, stroke, MI, BNP elevation, and ICU transfer
Secondary outcomes were each component of the composite outcome, hypotensive episodes (SBP < 100 mmHg), LOS, and discharge disposition
All outcomes were assessed following the initial 48 hours of admission and through hospital discharge.
Subgroup analysis:
There were 5 subgroup analyses to look at the outcomes of different subgroups interventions in intensive blood pressure management.
Patients older than 75 years
Patients with frailty
Patients with elevated outpatient BP > 140/90 mmHg
Patients with an SBP > 180 mmHg
Patients with cardiovascular disease
Additional analysis:
To evaluate the assumption that patients receiving more intensive BP treatment early in hospitalization would continue to receive intensive treatment, the number of antihypertensives and decrease in blood pressures were compared
Comparison of outcomes between patients who received IV antihypertensives versus oral antihypertensives
Statistical analysis:
A propensity score overlap weighting approach was used to negate the confounding by indication for treatment. Overlap weights, where treated patients were weighted for the probability of not receiving treatment and not treated patients were weighted for the probability of receiving treatment, were calculated.
Results
There were 176,050 hospitalized adults aged 65 years or older admitted with a non-cardiovascular condition between 10/01/2015 and 12/31/2017.
After applying exclusion criteria, there were 114,367 patients who qualified for the study, and 66,140 (57.8%) had 2 or more elevated blood pressures >= 140/90 mmHg. 14,064 (21.3%) patients received intensive treatment in the first 48 hours, and 52076 (78.7%) patients did not. Of those who received treatment, 2,504 patients (17.8%) received IV antihypertensives. The mean age of the group was 74.3, and 97.4% of the patients were men.
eFigure 1. Study flow chart from Anderson TS et al, JAMA Intern Med, 2023
Patients who received intensive treatment had higher systolic BPs prior to admission, with a mean of 140, and were more likely to have an SBP > 180 mmHg within the first 48 hours of admission.
Table 1: The cohort characteristics. From: Anderson TS et al, JAMA Intern Med, 2023
Patients in the intensive treatment group received a greater number of additional doses of medications for the remainder of their admission compared to those who did not (mean doses overall 6.1 [95% CI, 5.8-6.4] vs 1.6 [95% CI, 1.5–1.8]). Mean SBPs during the post 48-hour period were only slightly lower in the intensely treated group compared to the untreated group (138 mmHg [95% CI, 137.7–138.3 mmHg] vs 139.4 mmHg [95% CI, 139.2-139.6 mmHg]).
Figure 1. Change in systolic blood pressure by treatment group. From: Anderson TS et al, JAMA Intern Med, 2023
Patients who received intensive BP treatment were more likely to experience the primary composite outcome (1220 [8.7%] vs 3570 [6.9%]; weighted odds ratio [OR], 1.28; 95% CI, 1.18–1.39) and were more likely to experience the individual components of the composite outcome except for stroke and mortality. They were also more likely to experience hypotension with a SBP <100 mmHg. Intensely treated patients were less likely to get discharged home (OR=0.89).
Figure 2. Clinical outcomes of intensive outpatient treatment. From: Anderson TS et al, JAMA Intern Med, 2023
Patients who received IV antihypertensives had greater odds of the primary outcome than those who only received a PO antihypertensive [weighted OR, 1.90 (95%CI 1.65-2.19) vs weighted OR, 1.15 (95%CI 1.05-1.26)].
They also had increased odds of inpatient mortality compared to those who only received PO antihypertensives [weighted OR, 1.79 (95% CI, 1.26-2.53) vs weighted OR, 0.96 (95%CI, 0.76-1.21)].
Patients who received IV antihypertensives within 48 hours received a greater number of additional IV doses during the remainder of the hospitalization [1.4 (95 %CI 1.2-1.6) vs 0.1 (95%CI 0.1 to 0.2)].
Table 2. Clinical Outcomes of Intensive Inpatient Antihypertensive Treatment by Use of Intravenous vs Oral Antihypertensives. From: Anderson TS et al, JAMA Intern Med, 2023
For the 5 subgroups, intensive inpatient BP treatment was associated with higher rates of the primary outcome for each subgroup except for patients with a SBP >180mmHg in the first 48 hours.
Figure 3. Primary Composite Outcome of Intensive Inpatient Antihypertensive Treatment, Stratified by Age, Frailty, Outpatient Systolic Blood Pressure (BP), History of Cardiovascular Disease, and Maximum Systolic BP in the First 48 Hours. From: Anderson TS et al, JAMA Intern Med, 2023
Discussion
The study included about 20 percent of the patients who received intensive antihypertensive treatment per protocol. Similar to previous studies, additional blood pressure medications were associated with an increased risk of cardiac injury, acute kidney injury, and ICU transfers. The risk was higher still in patients who received intravenous antihypertensives. The result was uniform across subgroups of age, frailty, outpatient blood pressure control, and history of CVD.
This study’s findings were consistent with the previous study by Rastogi R et al. which also found an increased risk of myocardial injury and acute kidney injury with intensive antihypertensive therapy. Interestingly, this study didn’t find any increased risk of mortality and stroke with intensive antihypertensive treatment, as was seen in the study by Mohandas R et al. However, stroke and mortality were increased in patients who received IV antihypertensives.
Figure representing Inpatient Outcomes for Treated vs Untreated Patients by Treatment Route and SBP Interval. From: Rastogi R et al, JAMA Intern Med, 2020
The most logical reason for adverse outcomes for inpatients with intense blood pressure management is the sudden decrease in blood pressure of greater than 25 percent over a period of a few hours. This effect is even more prominent with IV antihypertensive medications, thus their association with even worse outcomes. Acute fluctuations of blood pressure can lead to watershed ischemia and end-organ damage.
None of the previous inpatient blood pressure management trials showed any positive signal with intense blood pressure control. However, it is essential to note that all studies to date have been observational due to the difficulty and expense of conducting a randomized control trial in this setting. Such trials would also be exceedingly difficult due to the heterogeneity of the indications for hospitalization and patient profiles.
This study once again supports the idea that asymptomatic inpatient hypertension does not necessarily need to be treated. Adequate confirmation of an elevated inpatient blood pressure is important, and measurement procedures (similar to those for outpatients) should be developed to allow for standardization and comparison. The authors suggest that an underlying etiology for an increase in blood pressure in an inpatient should be investigated, perhaps before the reflex action of treating it. This study supports the idea that the use of intravenous antihypertensives should be strongly discouraged in asymptomatic patients. The intensification of outpatient treatment based on inpatient blood pressure readings might be tempting, as many blood pressure readings are readily available and the patient is a captive audience while admitted to the hospital. Unfortunately, blood pressure management may not be related to the current admission or a patient's priority if they are acutely ill. In addition, changing medications can lead to confusion and an increased risk of overtreatment, potentially causing readmission.
Study’s limitations
There were several limitations to the study. As with priors, the present study is limited by its observational design. The study, conducted within the VHA healthcare system, which serves a predominantly elderly, Caucasian male population,may not be generalizable. Elderly populations are potentially more susceptible to adverse outcomes compared to younger patient groups. The study authors used biomarkers to identify cardiac injury rather than clinical events, and biomarkers were not completed on all patients, so it's possible patients may not have been captured as having an adverse event. Subjective symptoms of hypertensive emergencies, such as hypertensive encephalopathy, may not have been recognized as being related to hypertension, and therefore patients with these symptoms may not have been excluded. Additionally, patients may have been misclassified if they stopped any of their home antihypertensive therapies in the pre-hospitalization period.
Conclusion
Receipt of intensive inpatient antihypertensive treatment within the first 48 hours of admission was associated with a greater risk of adverse events (AKI, cardiac injury, ICU transfer), and even more so for those who received IV antihypertensives. The findings of this study do not support the use of intensive treatment for “elevated” blood pressure during hospital admission without evidence of end-organ damage. However, given the study's limitations, it's important for additional studies to capture a more diverse patient population to see if these findings apply to the general population.
““Caring about ourselves
This is our last dance
This is ourselves
Under pressure”
”
Summary by
Danielle Aliano
Internal Medicine PGY-3
Ochsner Medical Center, New Orleans, LA, USA
and
Jithu Kurian
Asst Professor - Nephrology
Pushpagiri Medical College, Thiruvalla, Kerala, India
NSMC Interns, Class of 2023
Reviewed by Brian Rifkin, Jade Teakell, Cristina Popa, and Swapnil Hiremath