Conference Coverage

Hospitalization risk twice as likely for veterans with mental illness


Key clinical point: In order to improve the health of veterans with mental illness, more efforts are needed to target alcohol use.

Major finding: Among veterans with mental illnesses, 7.1% were hospitalized, compared with 2.9% of their peers with no psychiatric or addiction diagnosis.

Study details: A database analysis of 952,252 veterans with a mental illness and 1,064,140 without a psychiatric or addiction diagnosis.

Disclosures: Dr. Moore reported having no financial disclosures.

Source: Moore et al. AAAP 2017. Paper session A1.



Veterans aged 60 and younger with mental illnesses were more than twice as likely to be hospitalized, compared with their peers who had no psychiatric or addiction diagnosis, according to results of a large VA database study.

“Our patients sit at the center of two public health crises,” David T. Moore, MD, PhD, said at the annual meeting and scientific symposium of the American Academy of Addiction Psychiatry. “One is the incredibly reduced life expectancy for adults with mental illnesses. There may be a 20-year reduced life expectancy. Their mortality rate for chronic medical conditions such as heart disease and COPD is increased by two- to fourfold, and is associated with greater hospitalization rates, longer lengths of stay, and increased readmission rates. The second part of this crisis is the incredible cost associated with medical hospitalizations. About 1 in every $20 in the entire U.S. economy goes toward inpatient medical hospitalization.”

Dr. David Moore

Dr. David Moore

In an effort to compare the risk of medical hospitalizations among veterans with and without mental health disorders, Dr. Moore, a psychiatrist at Yale University, New Haven, Conn., and his colleague, Robert Rosenheck, MD, identified 2,016,392 veterans under the age of 60 from the fiscal year 2012 Veterans Health Administration (VHA) databases. “We chose this group because it removes some confounding from nursing home visits, which also get coded in the VHA,” he explained. “It also removes some confounding from dementias and associated psychoses that could confuse our data.”

The final analysis included 952,252 veterans with a mental illness and 1,064,140 without a psychiatric or addiction diagnosis. Dr. Moore reported that among veterans with mental illness, 100,191 (7.1%) were hospitalized on a medical unit at some point during the study period, compared with only 31,759 (2.9%) of veterans with no psychiatric or addiction diagnosis. The Charlson Comorbidity Index was significantly increased in veterans with mental health diagnoses, compared with those who did not have mental health diagnoses.

“There was more tobacco use; they were much more likely to receive an opioid prescription; [and] they used more outpatient medical services, whether it be primary care visits or specialty care visits,” Dr. Moore said of the hospitalized veterans. “They are sicker, but they also use more outpatient medical services, suggesting that they do not lack access to adequate outpatient medical care.”

Next, the researchers performed a subset analysis of all veterans with any mental health diagnosis. Compared with those who were not hospitalized during the study period, those hospitalized were older (a mean age of 52 vs. 45 years, respectively), more likely to be homeless (21% vs. 12%; relative risk, 1.8), and receive a VA pension, which is correlated with poor functioning and disability (7.1% vs. 2.9%; RR, 2.4). The only psychiatric disorder correlated with correlated with medical hospitalization was personality disorder (6.3% vs. 3.7%; RR, 1.7). The researchers also observed that a higher proportion of hospitalized patients had an alcohol use disorder (34% vs. 23%; RR, 1.7) and drug use (31% vs. 17%; RR, 1.8). “The use of benzodiazepines had the greatest relative risk for medical hospitalizations,” Dr. Moore said.

In unadjusted analyses, veterans with the following diagnoses were at increased risk for hospitalization: drug use disorder (odds ratio, 4.58), alcohol use disorder (OR, 3.84), bipolar disorder (OR, 3.29), major depressive disorder (OR, 3.04), schizophrenia (OR, 2.98), and posttraumatic stress disorder (OR, 1.91).

After adjusting for other health factors in multiple regression, alcohol use disorder was the only psychiatric or addiction disorder strongly associated with medical hospitalizations (OR, 1.95). After accounting for sociodemographic characteristics, medical comorbidities, use of outpatient medical services, and alcohol use, the OR for medical hospitalizations among veterans with mental illness decreased from 2.52 to 1.24.

“It looks like a lot of the folks with drug use disorders who are being hospitalized may also have co-occurring alcohol use disorders,” Dr. Moore said. “That may partly account for their hospitalization risk.”

He concluded that the study’s overall findings “leave us with a lot of questions about what to do. The majority of patients who are hospitalized have a mental illness. Is this a setting where we should be engaging them and trying to connect them with outpatient services?”

Dr. Moore reported having no financial disclosures.

SOURCE: Moore et al. AAAP 2017. Paper session A1.

Next Article:

   Comments ()

Recommended for You

News & Commentary

Quizzes from MD-IQ

Research Summaries from ClinicalEdge