Overdose fatality and surveillance as a method for understanding mortality trends in homeless populations

"Healthy immigrant effect" among individuals experiencing homelessness in Spain?: Foreign-born individuals had higher average age at death in 15-year retrospective cohort study

BACKGROUND

Individuals experiencing homelessness (IEHs) suffer from severe health inequities. Place of origin is linked to health and mortality of IEHs. In the general population the "healthy immigrant effect" provides a health advantage to foreign-born people. This phenomenon has not been sufficiently studied among the IEH population. The objectives are to study morbidity, mortality, and age at death among IEHs in Spain, paying special attention to their origin (Spanish-born or foreign-born) and to examine correlates and predictors of age at death.

METHODS

Retrospective cohort study (observational study) of a 15-year period (2006-2020). We included 391 IEHs who had been attended at one of the city's public mental health, substance use disorder, primary health, or specialized social services. Subsequently, we noted which subjects died during the study period and analyzed the variables related to their age at death. We compared the results based on origin (Spanish-born vs. foreign-born) and fitted a multiple linear regression model to the data to establish predictors of an earlier age at death.

RESULTS

The mean age at death was 52.38 years. Spanish-born IEHs died on average almost nine years younger. The leading causes of death overall were suicide and drug-related disorders (cirrhosis, overdose, and chronic obstructive pulmonary disease [COPD]). The results of the linear regression showed that earlier death was linked to COPD (b = - 0.348), being Spanish-born (b = 0.324), substance use disorder [cocaine (b =-0.169), opiates (b =.-243), and alcohol (b =-0.199)], cardiovascular diseases (b = - 0.223), tuberculosis (b = - 0.163), high blood pressure (b =-0.203), criminal record (b =-0.167), and hepatitis C (b =-0.129). When we separated the causes of death for Spanish-born and foreign-born subjects, we found that the main predictors of death among Spanish-born IEHs were opiate use disorder (b =-0.675), COPD (b =-0.479), cocaine use disorder (b =-0.208), high blood pressure (b =-0.358), multiple drug use disorder (b =-0.365), cardiovascular disease (b =-0.306), dual pathology (b =-0.286), female gender (b =-0.181), personality disorder (b =-0.201), obesity (b =-0.123), tuberculosis (b =-0.120) and having a criminal record (b =-0.153). In contrast, the predictors of death among foreign-born IEHs were psychotic disorder (b =-0.134), tuberculosis (b =-0.132), and opiate (b =-0.119) or alcohol use disorder (b =-0.098).

CONCLUSIONS

IEHs die younger than the general population, often due to suicide and drug use. The healthy immigrant effect seems to hold in IEHs as well as in the general population.

Multiple substance use and blood pressure in women experiencing homelessness

Background

Substance use increases risk of cardiovascular events, particularly among women with additional risk factors like housing instability. While multiple substance use is common among unstably housed individuals, relationships between multiple substance use and cardiovascular risk factors like blood pressure are not well characterized.

Methods

We conducted a cohort study between 2016 and 2019 to examine associations between multiple substance use and blood pressure in women experiencing homelessness and unstable housing. Participants completed six monthly visits including vital sign assessment, interview, and blood draw to assess toxicology-confirmed substance use (e.g., cocaine, alcohol, opioids) and cardiovascular health. We used linear mixed models to evaluate the outcomes of systolic and diastolic blood pressure (SBP; DBP).

Results

Mean age was 51.6 years; 74 % were women of color. Prevalence of any substance use was 85 %; 63 % of participants used at least two substances at baseline. Adjusting for race, body mass index and cholesterol, cocaine was the only substance significantly associated with SBP (4.71 mmHg higher; 95 % CI 1.68, 7.74) and DBP (2.83 mmHg higher; 95 % CI 0.72, 4.94). Further analysis found no differences in SBP or DBP between those with concurrent use of other stimulants, depressants, or both with cocaine, compared to those who used cocaine only.

Conclusions

Cocaine was the only substance associated with higher SBP and DBP, even after accounting for simultaneous use of other substances. Along with interventions to address cocaine use, stimulant use screening during cardiovascular risk assessment and intensive blood pressure management may improve cardiovascular outcomes among women experiencing housing instability.

A Chronic Condition Disguised as an Acute Event: the Case for Re-thinking Stimulant Overdose Death

Recent reports indicate that stimulant-related deaths are increasing dramatically. People who die from acute stimulant toxicity have high rates of pre-existing cardiovascular disease (CVD), much of which is undiagnosed. Moreover, people who use stimulants with CVD often remain asymptomatic until presenting to an emergency department with an acute event. Prior research shows that symptoms of stimulant toxicity may occur on a regular basis, and that people who die from stimulant toxicity are older than those who die of opioid toxicity. Taken collectively, the existing evidence suggests that death from acute stimulant toxicity is often an outcome of long-term, cumulative exposure leading to cardiovascular dysfunction rather than acute intoxication. Strategies tailored to the distinct etiology of stimulant overdose are needed. We propose a three-part approach including (1) implementing stimulant use interventions that promote not only abstinence, but also use reduction, (2) treating ongoing stimulant use as a chronic cardiovascular condition, and (3) making stimulant toxicity interventions relevant to the populations most affected, which includes people outside of the traditional health-care system. In short, to reduce stimulant-related fatality, we need to transform our approach in ways that are tailored to address its natural history.

Multiple substance use, inflammation and cardiac stretch in women living with HIV

BACKGROUND

Cardiovascular disease (CVD) and heart failure (HF) are disproportionately high in people living with HIV and differ by sex. Few CVD-related studies focus on drug use, yet it is common in low-income women living with HIV (WLWH) and increases cardiac dysfunction.

SETTING

We recruited unsheltered and unstably housed WLWH from San Francisco community venues to participate in a six-month cohort study investigating linkages between drug use, inflammation, and cardiac dysfunction.

METHODS

Adjusting for CVD risk factors, co-infections, medications, and menopause, we examined the effects of toxicology-confirmed drug use and inflammation (C-reactive protein, sCD14, sCD163 and sTNFR2) on levels of NT-proBNP, a biomarker of cardiac stretch and HF.

RESULTS

Among 74 WLWH, the median age was 53 years and 45 % were Black. At baseline, 72 % of participants had hypertension. Substances used included tobacco (65 %), cannabis (53 %), cocaine (49 %), methamphetamine (31 %), alcohol (28 %), and opioids (20 %). Factors significantly associated with NT-proBNP included cannabis use (Adjusted Relative Effect [ARE]: -39.6 %) and sTNFR2 (ARE: 65.5 %). Adjusting for heart failure and restricting analyses to virally suppressed persons did not diminish effects appreciably. Cannabis use was not significantly associated with sTNFR2 and did not change the association between sTNFR2 and NT-proBNP.

CONCLUSIONS

Among polysubstance-using WLWH, NT-proBNP levels signaling cardiac stretch were positively associated with sTNFR2, but 40 % lower in people who used cannabis. Whether results suggest that cardiovascular pathways associated with cannabis use mitigate cardiac stress and dysfunction independent of inflammation in WLWH who use multiple substances merits further investigation.

Emergency department visits and trends related to cocaine, psychostimulants, and opioids in the United States, 2008-2018

BACKGROUND

Drug-related emergency department (ED) visits are escalating, especially for stimulant use (i.e., cocaine and psychostimulants such as methamphetamine). We sought to characterize rates, presentation, and management of ED visits related to cocaine and psychostimulant use, compared to opioid use, in the United States (US).

METHODS

We used 2008-2018 National Hospital Ambulatory Medical Care Survey data to identify a nationally representative sample of ED visits related to cocaine and psychostimulant use, with opioids as the comparator. To make visits mutually exclusive for analysis, we excluded visits related to 2 or more of the three possible drug categories. We estimated annual rate trends using unadjusted Poisson regression; described demographics, presenting concerns, and management; and determined associations between drug-type and presenting concerns (categorized as psychiatric, neurologic, cardiopulmonary, and drug toxicity/withdrawal) using logistic regression, adjusting for age, sex, race/ethnicity, and homelessness.

RESULTS

Cocaine-related ED visits did not significantly increase, while psychostimulant-related ED visits increased from 2008 to 2018 (2.2 visits per 10,000 population to 12.9 visits per 10,000 population; p < 0.001). Cocaine-related ED visits had higher usage of cardiac testing, while psychostimulant-related ED visits had higher usage of chemical restraints than opioid-related ED visits. Cocaine- and psychostimulant-related ED visits had greater odds of presenting with cardiopulmonary concerns (cocaine adjusted odds ratio [aOR] 2.95, 95% CI 1.70-5.13; psychostimulant aOR 2.46, 95% CI 1.42-4.26), while psychostimulant-related visits had greater odds of presenting with psychiatric concerns (aOR 2.69, 95% CI 1.83-3.95) and lower odds of presenting with drug toxicity/withdrawal concerns (aOR 0.47, 95%CI 0.30-0.73) compared to opioid-related ED visits.

CONCLUSION

Presentations for stimulant-related ED visits differ from opioid-related ED visits: compared to opioids, ED presentations related to cocaine and psychostimulants are less often identified as related to drug toxicity/withdrawal and more often require interventions to address acute cardiopulmonary and psychiatric complications.

Non-fatal stimulant overdose among homeless and unstably housed women in San Francisco, California

BACKGROUND

US rates of overdose deaths involving stimulants (e.g., cocaine or methamphetamine) have increased, but little is known about non-fatal stimulant overdoses, particularly among vulnerable populations. We characterized rates of non-fatal stimulant overdose identified outside of health care settings among women at high risk.

METHODS

Homeless and unstably housed women in San Francisco, California using stimulants were administered questionnaires on drug use and outcomes (stimulant overdose, health care utilization) monthly for six months. Based on pilot interviews, stimulant overdose during follow-up was defined as acute toxicity from stimulant use ("over-amping") resulting in "feeling sick, really scared, or like one's life may be in danger". Poisson regression estimated unadjusted incidence rate ratio (IRR) comparing participant characteristics.

RESULTS

We included 160 women (41% Black, 26% White, 15% Latina, median age 54 years) using crack cocaine (81%), methamphetamine (48%), and powdered cocaine (36%). Participants reported 67 non-fatal stimulant overdoses over 685 person-months of observation, a rate of 117.4 per 100 person-years (95% CI 85.8-160.5). Rates were higher among participants who were Latina vs. White (IRR 4.18 [1.60-10.94]), used methamphetamine (IRR 1.80 [0.96-3.38]), or used any stimulant daily/almost daily (IRR 2.63 [1.41-4.91]). Among women reporting stimulant overdose, 4% received emergency and 3% inpatient care for overdose of any drug.

CONCLUSIONS

Women in this setting, particularly those who used stimulants frequently or used methamphetamine, experienced high non-fatal stimulant overdose and rarely received health care for these events. Efforts should be made to increase awareness and reduce harms of stimulant toxicity in vulnerable populations.

Using a "Big Events" framework to understand emergency department use among women experiencing homelessness or housing instability in San Francisco during the COVID-19 pandemic

Background

The COVID-19 pandemic created a major public health crisis that disrupted economic systems, social networks and individual behaviors, which led to changes in patterns of health care use. Factors associated with emergency department (ED) visits during the pandemic among especially high-risk individuals are unknown. We used a “Big Events” approach, which considers major disruptions that create social instability, to investigate ED use in people experiencing homelessness or housing instability, many of whom use drugs.

Methods

Between July and December 2020, we conducted a community-based San Francisco study to compare homeless and unstably housed (HUH) women who did and did not use an ED during the first 10 months of the pandemic.

Results

Among 128 study participants, 34% had ≥1 ED visit during the pandemic. In adjusted analysis, factors significantly associated with ED use included experiencing homelessness, cocaine use and increased difficulties receiving drug use treatment during the pandemic.

Conclusion

These findings build on the “Big Events” approach to considering risk pathways among people who use drugs. They suggest the importance of ensuring access to housing and low-barrier non-COVID health services, including drug treatment, alongside crisis management activities, to reduce the health impacts of public health crises.

Homelessness and Veteran Status in Relation to Nonfatal and Fatal Opioid Overdose in Massachusetts

BACKGROUND

Compared with non-Veterans, Veterans are at higher risk of experiencing homelessness, which is associated with opioid overdose.

OBJECTIVE

To understand how homelessness and Veteran status are related to risks of nonfatal and fatal opioid overdose in Massachusetts.

DESIGN

A cross-sectional study.

PARTICIPANTS

All residents aged 18 years and older during 2011-2015 in the Massachusetts Department of Public Health's Data Warehouse (Veterans: n=144,263; non-Veterans: n=6,112,340). A total of 40,036 individuals had a record of homelessness, including 1307 Veterans and 38,729 non-Veterans.

MAIN MEASURES

The main independent variables were homelessness and Veteran status. Outcomes included nonfatal and fatal opioid overdose.

RESULTS

A higher proportion of Veterans with a record of homelessness were older than 45 years (77% vs. 48%), male (80% vs. 62%), or receiving high-dose opioid therapy (23% vs. 15%) compared with non-Veterans. The rates of nonfatal and fatal opioid overdose in Massachusetts were 85 and 16 per 100,000 residents, respectively. Among individuals with a record of homelessness, these rates increased 31-fold to 2609 and 19-fold to 300 per 100,000 residents. Homelessness and Veteran status were independently associated with higher odds of nonfatal and fatal opioid overdose. There was a significant interaction between homelessness and Veteran status in their effects on risk of fatal overdose.

CONCLUSIONS

Both homelessness and Veteran status were associated with a higher risk of fatal opioid overdoses. An understanding of health care utilization patterns can help identify treatment access points to improve patient safety among vulnerable individuals both in the Veteran population and among those experiencing homelessness.

Impact of polysubstance use on high-sensitivity cardiac troponin I over time in homeless and unstably housed women

INTRODUCTION

The use of controlled substances like cocaine increases the risk of cardiovascular disease (CVD) and myocardial infarction (MI). However, outside of alcohol and tobacco, substance use is not included in CVD risk assessment tools. We identified the effects of using multiple substances (nicotine/cotinine, cannabis, alcohol, cocaine, methamphetamine, heroin and other opioids) on cardiac injury measured by high-sensitivity troponin (hsTnI) in homeless and unstably housed women.

METHODS

We recruited 245 homeless and unstably housed women from shelters, free meal programs and street encampments. Participants completed six monthly study visits. Adjusting for traditional CVD risk factors, we examined longitudinal associations between substance use and hsTnI.

RESULTS

Median participant age was 53 years and 74 % were ethnic minority women. At baseline, 76 % of participants had hypertension, 31 % were HIV-positive, 8% had a history of a prior MI and 12 % of prior stroke. The most commonly used substances were cotinine/nicotine (80 %), cannabis (68 %) and cocaine (66 %). HsTnI exceeding the 99th percentile (14.7 ng/L) - a level high enough to signal possible MI - was observed in 14 participants during >1 study visit (6%). In adjusted analysis, cocaethylene and fentanyl were significantly associated with higher hsTnI levels.

CONCLUSIONS

Fentanyl use and the co-use of cocaine and alcohol are associated with myocardial injury, suggesting that the use of these substances may act as long-term cardiac insults. Whether risk counseling on these specific substances and/or including their use in CVD risk stratification would improve CVD outcomes in populations where substance use is high merits further investigation.

Drug overdose mortality among residents of single room occupancy buildings in San Francisco, California, 2010-2017

BACKGROUND

Single room occupancy (SRO) buildings, also known as residential hotels, are a form of affordable housing common to cities in North America, and residents of these buildings face elevated rates of substance use, physical and mental multimorbidity, and mortality. Identifying distinct populations at greater risk of overdose death is crucial to the planning of interventions aiming to reduce drug-related mortality, yet no studies have assessed the population burden of overdose mortality among SRO residents. The present study quantifies and characterizes drug overdose mortality among residents of SRO buildings in a large U.S. city.

METHODS

We used mortality records and a database of SRO buildings to calculate rate ratios comparing overdose mortality due to opioids, cocaine, and methamphetamine among SRO residents and non-SRO residents in San Francisco, CA 2010-2017 and assessed bivariate differences in decedent and death location characteristics between SRO resident and other overdose decedents.

RESULTS

There were 1,551 overdose deaths during the study period, with an overall rate of 21.3 per 100,000 residents (95%CI = 20.2-22.6). The rate among SRO residents (278.7, 95%CI = 252.9-306.5) was 19.3 (95%CI = 17.1-21.7) times that of non-SRO residents (21.3, 95%CI = 20.2-22.6). An additional 79 (5%) deaths among non-residents occurred in SRO buildings, and 86% of SRO resident decedents died at home compared to 64% of non-SRO residents (p < 0.05).

CONCLUSIONS

Overdose mortality was substantially higher among SRO residents, who were also more likely to die from overdose at home, which highlights the need for resources and targeted interventions directed towards residents of SRO buildings.

Associations between medical cannabis and other drug use among unstably housed women

BACKGROUND

Several studies suggest that U.S. state-level legalization of cannabis for medical purposes may be associated with reductions in opioid use; yet its relationship with stimulant use, particularly in high-risk populations like unstably housed women, has received less attention. The purpose of this study was to estimate independent associations between medical and non-medical use of cannabis and use of stimulants and opioids among unstably housed women.

METHODS

Cross-sectional data were analyzed from 245 women in the SHADOW study, a community based cohort in San Francisco, CA, in which HIV+ women were oversampled (126 HIV+ and 119 HIV-).

RESULTS

Compared to no cannabis use in the past 6 months (51%), non-medical cannabis use (28%) was associated with a higher adjusted odds of using stimulants (Adjusted Odds Ratio [AOR]=4.34, 95% confidence interval [CI]: 2.17-8.70) and opioids (AOR=3.81, 95% CI: 1.78-8.15). Compared to no cannabis use, medical cannabis use (21%) was not significantly associated with stimulant or opioid use. Compared to non-medical cannabis use, however, medical cannabis use was associated with lower adjusted odds of using stimulants (AOR=0.42, 95% CI: 0.18-0.96). These associations were not modified by HIV status.

CONCLUSIONS

Associations between use of cannabis and "street drugs" depend on whether the cannabis is obtained through a medical context. Interventions, research, and policy considering the influence of cannabis on the use of other drugs may benefit by distinguishing between medical and non-medical cannabis use.

Higher prevalence of detectable troponin I among cocaine-users without known cardiovascular disease

BACKGROUND

While cocaine use is an established risk factor for acute cardiovascular complications, associations between cocaine use and markers of cardiac injury outside of acute hospital presentation remain poorly characterized. We leveraged advances in cardiac troponin (cTnI) testing to assess low but clinically meaningful levels of cardiac injury among cocaine users and non-users.

METHODS

We conducted a case control study comparing cTnI levels by the presence of cocaine among patients presenting for non-cardiac care in an urban safety net hospital. Samples were chosen sequentially among those for which urine drug screens were ordered by providers hospital-wide.

RESULTS

During 2015, 14% of all hospital drug screens ordered were cocaine-positive. Among unique persons providing cocaine-positive (N=100) and cocaine-negative (N=100) samples, 37% were female, 45% were African-American and the median age was 51. Detectable cTnI (> 0.02ng/mL) was observed in 21 samples (11%). It was more common in subjects using cocaine (Adjusted OR=2.81; 95% CI=1.03-7.65), but not other drugs. Moreover, there was a significant correlation between concentrations of cTnI and the cocaine metabolite, benzoylecgonine (Spearman Correlation=0.34, p<0.01).

CONCLUSIONS

Among urban safety net hospital patients, 11% had detectable cTnI, and cTnI concentration was significantly correlated with benzoylecgonine concentration. While these preliminary results require additional confirmation, they suggest the potential utility of considering cocaine use as more than just an episodic exposure leading to acute cardiac events. The consideration of cocaine use as an ongoing chronic exposure leading to subclinical cardiac injury may improve risk-stratification and patient outcomes in populations where cocaine use is high.

Characteristics of Homeless Adults Who Died of Drug Overdose: A Retrospective Record Review

Drug overdose is a major cause of death among homeless people, but little is known about the characteristics of homeless overdose decedents. We conducted a retrospective record review of 219 adult patients of Boston Health Care for the Homeless Program (BHCHP) who died of drug overdose in 2003-2008. We assessed the substances implicated in overdose and the health and service use characteristics of decedents prior to death. Eighty-one percent of overdose deaths involved opioids and 40% involved multiple drugs. Problem substance use (85%), psychiatric illness (61%), and chronic pain (45%) were common, and 32% had documentation of all three. Half were well-connected to BHCHP, and 35% had a clinic visit within 90 days of death. The complex health histories and frequent health care contacts of homeless drug overdose decedents suggest that clinical facilities may be an important frontline venue for overdose education, naloxone distribution, and integrated substance use treatment programming.

Risk factors for stimulant use among homeless and unstably housed adult women

BACKGROUND

One of the most common causes of death among homeless and unstably housed women is acute intoxication where cocaine is present. While correlates of stimulant use have been determined in prior research, few studies have assessed risk factors of use specifically in this high-risk population.

METHODS

We sampled biological women with a history of housing instability from community-based venues to participate in a cohort study. Baseline and 6-month follow-up data were used to determine the relative risk of stimulant use (crack cocaine, powder cocaine or methamphetamine) among individuals who did not use at baseline.

RESULTS

Among 260 study participants, the median age was 47 years, 70% were women of color; 47% reported having unmet subsistence needs and 53% reported abstinence from stimulants at baseline. In analyses adjusting for baseline sociodemographics and drug treatment, the risk of using stimulants within 6 months was significantly higher among women who reported recent sexual violence (Adjusted Relative Risk [ARR]=4.31; 95% CI:1.97-9.45), sleeping in a shelter or public place (ARR=2.75; 95% CI:1.15-6.57), and using unprescribed opioid analgesics (ARR=2.54; 95% CI:1.01-6.38).

CONCLUSION

We found that almost half of homeless and unstably housed women used stimulants at baseline and 14% of those who did not use began within 6 months. Addressing homelessness and sexual violence is critical to reduce stimulant use among impoverished women.

Tobacco-, alcohol-, and drug-attributable deaths and their contribution to mortality disparities in a cohort of homeless adults in Boston

OBJECTIVES

We quantified tobacco-, alcohol-, and drug-attributable deaths and their contribution to mortality disparities among homeless adults.

METHODS

We ascertained causes of death among 28 033 adults seen at the Boston Health Care for the Homeless Program in 2003 to 2008. We calculated population-attributable fractions to estimate the proportion of deaths attributable to tobacco, alcohol, or drug use. We compared attributable mortality rates with those for Massachusetts adults using rate ratios and differences.

RESULTS

Of 1302 deaths, 236 were tobacco-attributable, 215 were alcohol-attributable, and 286 were drug-attributable. Fifty-two percent of deaths were attributable to any of these substances. In comparison with Massachusetts adults, tobacco-attributable mortality rates were 3 to 5 times higher, alcohol-attributable mortality rates were 6 to 10 times higher, and drug-attributable mortality rates were 8 to 17 times higher. Disparities in substance-attributable deaths accounted for 57% of the all-cause mortality gap between the homeless cohort and Massachusetts adults.

CONCLUSIONS

In this clinic-based cohort of homeless adults, over half of all deaths were substance-attributable, but this did not fully explain the mortality disparity with the general population. Interventions should address both addiction and non-addiction sources of excess mortality.

Predictors of Frequent Emergency Room Visits among a Homeless Population

Background

Homelessness, HIV, and substance use are interwoven problems. Furthermore, homeless individuals are frequent users of emergency services. The main purpose of this study was to identify risk factors for frequent emergency room (ER) visits and to examine the effects of housing status and HIV serostatus on ER utilization. The second purpose was to identify risk factors for frequent ER visits in patients with a history of illicit drug use.

Methods

A retrospective analysis was performed on 412 patients enrolled in a Boston-based health care for the homeless program (HCH). This study population was selected as a 2:1 HIV seronegative versus HIV seropositive match based on age, sex, and housing status. A subgroup analysis was performed on 287 patients with history of illicit drug use. Chart data were analyzed to compare demographics, health characteristics, and health service utilization. Results were stratified by housing status. Logistic models using generalized estimating equations were used to predict frequent ER visits.

Results

In homeless patients, hepatitis C was the only predictor of frequent ER visits (OR 4.49, p<0.01). HIV seropositivity was not predictive of frequent ER visits. In patients with history of illicit drug use, mental health (OR 2.53, 95% CI 1.07–5.95) and hepatitis C (OR 2.85, 95% CI 1.37–5.93) were predictors of frequent ER use. HIV seropositivity did not predict ER use (OR 0.45, 95% CI 0.21 – 0.97).

Conclusions

In a HCH population, hepatitis C predicted frequent ER visits in homeless patients. HIV seropositivity did not predict frequent ER visits, likely because HIV seropositive HCH patients are engaged in care. In patients with history of illicit drug use, hepatitis C and mental health disorders predicted frequent ER visits. Supportive housing for patients with mental health disorders and hepatitis C may help prevent unnecessary ER visits in this population.