Professional Documents
Culture Documents
Tgs Jurnal 1
Tgs Jurnal 1
Hypertension is an important modifiable risk factor for car- of end-organ damage and do not require immediate work
diovascular disease. In the United States, nearly 1 out of 3 up and aggressive treatment.4 These patients with severe
adults have elevated blood pressure, with the highest rates asymptomatic hypertension comprise 5% of emergency
among African Americans. Within adults with hyperten- department visits and can be managed with oral antihyper-
sion, 82% are aware of their disease and only 53% have their tensive medications, usually as outpatients.5 Hypertensive
blood pressure controlled to target levels.1,2 Hypertension emergency (HTNE) or acute severe hypertension previously
remains the commonest encountered problem in primary known as malignant hypertension is characterized by an
care and is the most important modifiable risk factor when accelerated increase in blood pressure secondary to increase
it comes to prevention of myocardial infarction, stroke, and in catecholamines, sympathetic nervous system activity,
renal failure.3 endothelial dysfunction, renin–angiotensin system activa-
A majority of patients with severely elevated blood pres- tion, or acute stress and is linked with acute end-organ dam-
sure (blood pressure ≥ 180/120 mm Hg) have no evidence age.6 Evidence of cardiac injury or infarction, pulmonary
edema, acute renal failure, stroke, and/or aortic dissection in chronic pulmonary disease, peripheral vascular disease, and
the setting of acute rise in blood pressure are considered as secondary hypertension), length of stay, hospital charges, All
end-organ damage.7 Fortunately, the reported incidence of Patient Refined Diagnosis Related Groups (APRDRG) sub-
HTNE is 1 to 2 cases per million per year. classes for mortality and severity of illness risk score, proce-
Approximately 1% of patients with known hyperten- dures such as echocardiography, coronary revascularization
sion develop HTNE during their lifetime and a majority of [either percutaneous coronary intervention or bypass sur-
HTNE cases are considered secondary to inadequate treat- gery (CABG)], and acute end-organ complications/symp-
ment.8 In a case series of 100 patients admitted for HTNE, toms associated with HTNE [acute myocardial infarction,
50% of the patients with pre-existing hypertension had aortic dissection, acute chest pain, acute renal failure, acute
stopped taking their antihypertensive medications more heart failure including acute pulmonary edema, acute neu-
than a month prior to admission.9 A nationwide study in rological symptoms (hypertensive encephalopathy, confu-
Italy showed that 8% of visits for HTNE had no prior docu- sion, altered mental status), and stroke or transient ischemic
mentation of hypertension.10 With the advent of safe and attack (TIA)]. We defined severity of comorbid conditions
improved antihypertensive therapies, outcomes among using Deyo modification of Charlson comorbidity index
0.3% per year by 2012. By the year 2007, there was a 50% comorbidities such as diabetes, coronary artery disease, and
relative reduction in rate of mortality which was sustained secondary hypertension.
thereafter. We also compared mortality rates monthly, which When we studied baseline socioeconomic status, we
did not show any statistical difference (P = 0.43). noticed that patients within the lowest income quartile for
Additionally, we analyzed the annual trends in APRDRG ZIP code had highest admission rate. Patients in the lowest
subclasses for mortality and severity of illness risk score ≥3 25th percentile for income accounted for nearly 37% of total
in addition to the CCI. We found that there was a statisti- admissions, but experienced a similar in-hospital mortality
cally significant uptrend in the proportion of patients with rate compared to rest of the population (0.47% vs. 0.49%;
the higher grades of APRDRG subclasses for mortality risk P = 0.53). Forty-nine percent of all admitted patients had
≥3 (10.2% vs. 17.2% in 2002 and 2012, respectively), sever- Medicare as primary insurer.
ity of illness score ≥3 (24.6% vs. 32.3% in 2002 and 2012,
respectively), and the mean CCI (2.95 vs. 3.35 in 2002 and Procedures
2012, respectively).
The use of echocardiogram was higher in patients who
Baseline characteristics died than survived (7.0% vs. 3.8%; P < 0.001). There was no
significant difference in the use of coronary angiogram in
Patients were older (67.9 ± 16.6 years vs. 58.9 ± 17.1 years; the 2 groups (5.9% vs. 5.0%; P = 0.33); however, a higher
P < 0.001) within the group that experienced in-hospital proportion of patients received revascularization within the
mortality compared to those who were alive at discharge. group with in-hospital death (2.2% vs. 0.5%; P < 0.001).
Most of the patients were admitted to urban nonteaching
hospitals and regional variations in HTNE admissions were Outcomes and association with acute end-organ
seen throughout the duration of the study with a consist- complications
ently larger proportion of admissions being from the South
(52.3%). The incidence of all studied acute end-organ complica-
Table 1 represents the baseline differences between tions was consistently higher within the group with in-
patients who experienced in-hospital mortality and those hospital mortality (Figure 2). Patients who died were more
who survived. Similar to age, the length of stay (11.7 days than 7 times as likely to suffer from acute cardiorespiratory
vs. 3.8 days; P < 0.001), and hospital charges ($88,899 vs. failure compared to the patients who survived (66.2 vs. 8.9;
$23,701; P < 0.001) were significantly higher within the for- P < 0.001). Single-organ damage was reported in 85.7% of the
mer group. Among those who experienced in-hospital mor- patients who died compared to 28.6% in the group that sur-
tality, a greater proportion of patients had peripheral arterial vived. More than half (56.5%) of the patients within the for-
disease (19% vs. 9.8%; P < 0.001), chronic pulmonary disease mer group had 2 or more reported end-organ complications.
(22% vs. 15.1%; P < 0.001), APRDRG mortality risk score, Using multivariate logistic regression analysis (Table 2),
APRDRG severity score, number of baseline chronic condi- end-organ related complications such as acute cardiores-
tions (7.9 ± 2.9 vs. 5.8 ± 2.7; P < 0.001), and CCI (3.6 ± 2.0 piratory failure [adjusted odds ratio (OR), 15.8; 95% con-
vs. 2. 0 ± 1.8; P < 0.001). Both the study groups were similar fidence interval (CI), 13.2–18.9], stroke or TIA (adjusted
when it came to race or gender, weekend admission rate, and OR, 7.9; 95% CI, 6.3–9.9), chest pain (adjusted OR, 5.9; 95%
Table 1. Differences in baseline characteristics, risk assessment scores, and end-organ involvement within the cohort
Abbreviations: APRDRG, All Patients Refined Diagnosis Related Groups; TIA, transient ischemic attack.
CI, 4.4–7.7), aortic dissection (adjusted OR, 5.9; 95% CI, predictive of higher in-hospital mortality. Patients who met
2.8–12.4) were most predictive of higher in-hospital mortal- the primary endpoint of in-hospital death had a significantly
ity. Among other risk factors, acute myocardial infarction, higher rate of end-organ complications, with almost 86% of
acute renal failure, presence of neurological symptoms, and the patients experiencing at least single-organ complications.
age were also predictive of higher mortality. Sex and African A study by Polgreen et al. examined the rising trend for
American race were not predictive of in-hospital mortality more severe hypertension-related diagnoses following 2007
(P > 0.05). within NIS, and hypothesized that there was a substantial
shift in the assigning of more severe diagnoses, less mild
DISCUSSION diagnoses, likely upcoding, and changes in billing practices.15
Alterations in coding practices have also been described in
The primary finding of this manuscript was that in a large other studies, some of which are influenced by reimburse-
multi-institutional observational cohort within the United ment practices, and payment policies put forward by the
States, there was a consistent increase in HTNE admissions Centers for Medicare and Medicaid Services.16,17 There was
with marked reduction of in-hospital mortality from 2002 to a trend toward increase in baseline comorbidity and risk
2012. Within the in-hospital mortality group were older, had scores during the same period among the HTNE patients in
longer hospital stays, higher cost of stay, more tests/proce- our study. We propose that finding may be in part secondary
dures, and a higher overall prevalence of baseline comorbid- to upcoding of diagnoses, but improvements in care during
ities such as peripheral arterial disease, chronic pulmonary this period are also likely to have played significant role in
disease, and clinical risk scores. Presence of acute cardi- the decreased mortality.
orespiratory failure, stroke/TIA, acute chest pain, aortic A recent study on HTNE patients reported a mortal-
dissection, acute myocardial infarction, acute renal failure, ity rate of 4.6%, much higher than reported in our study.18
neurological symptoms, and older age were all significantly However, they exclusively analyzed patients admitted to
the coronary care unit, thus inducing a selection bias and failure, acute renal failure, acute myocardial infarction,
including patients requiring critical inpatient care and had acute stroke or TIA, aortic dissection, and the presence of
a sample size of only 538 HTNE patients. Overall rate of in- hypertensive neurologic symptoms or acute chest pain as
hospital mortality in our study was more comparable to the independent factors affecting mortality. We did not find any
mortality rates seen with hypertensive urgency of 0.8% (126 significant gender-based variation in mortality, though has
patients) in the same study. A US registry consisting of 1,588 been reported in prior published literature. Among patients
patients admitted with acute severe hypertension requiring who died, 66.2% had acute heart failure, followed by renal
intravenous antihypertensives found patients to have an even failure (41.4%), neurological symptoms (21.3%), stroke/TIA
higher hospital mortality of 6.9%.11 When we separately ana- (19.2%), and acute myocardial infarction (16.7%). Not all
lyzed in-hospital mortality among our patients who had one HTNE is created equal. Patients developing acute cardiores-
or more of the studied complications, in-hospital mortal- piratory failure, stroke or TIA, and aortic dissection suffer-
ity rate was 1.4%. The relatively lower mortality rates (with ing from the worst outcomes.
reduction in mortality trends) were likely due to the assign- Additional prognostic information was also obtained by
ment of a diagnosis of HTNE among patients with lesser using risk scores within the NIS database. APRDRG is a clas-
degrees of hypertensive disease with relatively benign pres- sification system that classifies patients according to their
entation explained above. This can been seen in the relatively reason of admission, severity of illness, and risk of mortality
lower rate of end-organ damage in the group that survived and has been validated when it comes to predicting hospital
within our study, with acute renal failure being reported in outcomes.21 Patients with in-hospital mortality were more
half of those. The study by Leiba et al. analyzed 1-year HTNE likely to have a greater proportion of patients in the more
mortality rates at a large scale referral center in Israel over severe subclasses for both risk scores. Patient with in-hos-
the course of 20 years and found a significant reduction in pital mortality also had a higher CCI, a known prognostic
mortality rates during this time, which may be attributed to marker for higher hospital mortality among other factors.22
increased awareness of the worse outcomes associated with The incidence and mortality related to several cardio-
true diagnosis leading to prompt, efficient, and improved vascular emergencies especially myocardial infarction has
antihypertensive care.19 Interestingly, they reviewed 306 a clear tendency to increase during the winter months.23,24
patient files as part of their study and found that only 46% of Deshmukh et al. analyzed the 2000–2007 NIS database
the cases had a true diagnosis. and reported seasonal variations in the incidence of HTNE
To date, large scale data on specific predictors for mor- admissions, with a spike during the winter months. We ana-
tality in HTNE patients is missing. HTNE has been shown lyzed the trends in seasonal mortality and found no signifi-
to have variable presentations when it comes to end-organ cant difference in rates mortality throughout the year.
involvement. According to one 1996 study, cerebral infarc- There are some limitations to this study. The study is based
tion or hemorrhage (28.5%), acute pulmonary edema (23%), on administrative data. No charts were available and we did
and hypertensive encephalopathy (16%) were the common- not study out of hospital deaths. We are heavily dependent
est types of end-organ damage.11 Pacheco et al. found acute on coders, which may have wide variation and potential for
coronary syndrome (59.5%) and acute decompensated heart unrecognized miscoding. There is a possibility of selection
failure (25.2%) to be predominant in their assessment of bias and residual measured and unmeasured confounding
coronary care unit patients.10 Another study followed 297 factors, as this is a retrospective observational study. We
HTNE patients for a median of 30 months and found that were unable to assess beyond hospital stay or report upon
diabetes mellitus, acute left ventricular failure, stroke, and treatment strategies used. We hope that the large number of
renal impairment were independent predictors of mortal- database studied eliminates these limitations. We propose
ity.20 In our study, we identified age, acute cardiorespiratory incremental improvement in care during the time of the
study but are unable to evaluate the treatment strategies that 6. Lagi A, Cencetti S. Hypertensive emergencies: a new clinical approach.
may have been responsible. Our findings need validation in Clin Hypertens 2015; 21:20.
7. Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA,
other large registries or observational studies, as it would Izzo JL Jr, Jones DW, Materson BJ, Oparil S, Wright JT Jr, Roccella
enable us to risk stratify patients with HTNE and influence EJ; Joint National Committee on Prevention, Detection, Evaluation,
management. We strongly believe prospectively conducted and Treatment of High Blood Pressure. National Heart, Lung, and
studies with stringent criteria for diagnosis of HTNE are Blood Institute; National High Blood Pressure Education Program
Coordinating Committee. Seventh report of the Joint National
urgently needed to study the impact of end-organ complica- Committee on Prevention, Detection, Evaluation, and Treatment of
tions and specific treatment strategies on outcomes. High Blood Pressure. Hypertension 2003; 42:1206–1252.
8. Vidt DG. Current concepts in treatment of hypertensive emergencies.
Am Heart J 1986; 111:220–225.
9. Bennett NM, Shea S. Hypertensive emergency: case criteria, sociode-
SUPPLEMENTARY MATERIAL mographic profile, and previous care of 100 cases. Am J Public Health
1988; 78:636–640.
Supplementary data are available at American Journal of 10. Zampaglione B, Pascale C, Marchisio M, Cavallo-Perin P. Hypertensive
Hypertension online. urgencies and emergencies: prevalence and clinical presentation.