Skip to main content
NCBI home page
American Journal of Public Health logoLink to American Journal of Public Health
. 2012 Jun;102(6):1145–1153. doi: 10.2105/AJPH.2011.300598

Payment Source and Emergency Management of Deliberate Self-Harm

Steven C Marcus 1, Jeffrey A Bridge 1, Mark Olfson 1
PMCID: PMC3483957  PMID: 22515853

Abstract

Objectives. We investigated whether health insurance type (private vs Medicaid) influences the delivery of acute mental health care to patients with deliberate self-harm.

Methods. Using National Medicaid Analytic Extract Files (2006) and MarketScan Research Databases (2005–2007), we analyzed claims focusing on emergency episodes of deliberate self-harm of Medicaid- (n = 8228) and privately (n = 2352) insured adults. We analyzed emergency department mental health assessments and outpatient mental health visits in the 30 days following the emergency visit for discharged patients.

Results. Medicaid-insured patients were more likely to be discharged (62.7%), and among discharged patients they were less likely to receive a mental health assessment in the emergency department (47.8%) and more likely to receive follow-up outpatient mental health care (52.9%) than were privately insured patients (46.9%, 57.3%, and 41.2%, respectively).

Conclusions. Acute emergency management of deliberate self-harm is less intensive for Medicaid- than for privately insured patients, although discharged Medicaid-insured patients are more likely to receive follow-up care. Programmatic reforms are needed to improve access to emergency mental health services, especially in hospitals that serve substantial numbers of Medicaid-insured patients.

In several medical contexts, insurance status affects access and quality of care.1,2 In relation to privately insured patients, Medicaid-insured patients are generally less likely to receive care, and the care they receive less often meets guideline-based standards.3,4 At a time of intense pressure to rein in Medicaid costs, the risks of insurance-related disparities in access and quality of care are especially great. Despite keen interest among health care policymakers, however, little information exists in the peer-reviewed literature on the comparative access and quality of mental health care provided to Medicaid- and commercially insured patients.

The emergency department is an important setting in which to evaluate mental health care of Medicaid-insured and privately insured adults. In the management of deliberate self-harm, emergency departments have an opportunity to provide potentially lifesaving services. A key goal involves improving the recognition and treatment of psychiatric disorders in high-risk patients.5 One particularly high-risk group includes patients who go to emergency departments pursuant to self-harm, which may or may not involve suicidal intent.6,7 During the first year following emergency treatment of deliberate self-harm, suicide risk increases 30- to 130-fold.8–10 Psychiatric disorders, though exceedingly prevalent among adults with deliberate self-harm,11,12 are not always recognized or treated. Because of strong connections between psychiatric disorder, self-harm, and suicide, the National Institute for Clinical Excellence recommends that all individuals attending emergency departments with deliberate self-harm receive a mental health evaluation before emergency department discharge.13 Mental health assessment provides opportunities to evaluate aggression, impulsivity, hopelessness, and other symptoms that bear on suicide risk.14

The quality of mental health care in emergency departments varies. In 1 statewide survey, most emergency departments did not have a mental health professional to evaluate and treat patients with self-harm.15 Only about one half of emergency department patients in the Medicaid program who come in with deliberate self-harm and are discharged to the community receive a mental health assessment.16

We compared the mental health care received by adult Medicaid and privately insured beneficiaries who came to emergency departments with deliberate self-harm. We first compared these 2 groups with respect to the likelihood of hospital admission. Among those who were discharged to the community, we assessed the probability of receiving a mental health assessment in the emergency department and of receiving outpatient mental health treatment within the following 30 days. We hypothesized that Medicaid-insured patients with self-harm would be less likely than would their privately insured counterparts to be admitted to the hospital and that Medicaid-insured patients who were discharged to the community would be less likely than would privately insured patients to receive an emergency mental health assessment and follow-up outpatient mental health care.

METHODS

We analyzed data from the 2006 Medicaid Analytic Extract files from all 50 US states and the MarketScan Research Databases (2005–2007). MarketScan data are collected directly from more than 150 large employers, who offer 1 or more of 80 private health plans in 8 states.17 Medicaid is a state and federal program that pays for medical assistance for approximately 49 million individuals with low incomes and limited resources.18 Within federal guidelines, each state establishes eligibility standards, selects the range of services, and sets payment rates.19 Private health insurance plans cover approximately 196 million Americans under a range of benefit designs, cost-sharing arrangements, and provider reimbursement schemes.20 Individual emergency department charges paid to hospitals tend to be higher for privately insured visits than for Medicaid-insured visits.21

We derived supplementary county-level information on poverty, income, and population density from the Area Resource File.22

Sample Selection

We followed patients aged 21 to 64 years through claims histories for 60 days before and 30 days after each emergency department visit for deliberate self-harm (International Classification of Diseases [ICD]-9-CM:E950-E959). Because of the focus on connections to outpatient mental health care, we excluded patients if they received services in an institutional residential setting, such as a prison, group home, or residential care facility, during the follow-up period. Patients could contribute more than 1 deliberate self-harm–related emergency department treatment episode to the analysis. A total of 9539 patients contributed 10 578 treatment episodes.

Dependent Variables

The 3 dependent variables were inpatient admission, mental health assessment within the emergency department, and follow-up outpatient mental health care within 30 days of emergency department discharge. We limited our analysis of the latter 2 variables to patients who were discharged from the emergency department to the community. We defined mental health assessments as the occurrence of 1 or more claims with Current Procedural Terminology codes for psychiatric evaluation and diagnostic interviewing (90801, 90802, 90885), psychological and neuropsychological assessment and testing (96101–96103, 96118–96120), or diagnosis of a mental disorder (ICD-9-CM 290–319) during the emergency department visit. We defined follow-up outpatient mental health care as any reimbursed outpatient mental health service within 30 days of discharge. A combination of place of service and diagnosis codes defined outpatient mental health services. We selected a 30-day period according to the Healthcare Effectiveness Data and Information Set quality measures for follow-up after hospitalization for mental illness.23

Independent Variables

Independent variables included patient level and local resource variables. Patient level variables included age (21–34, 35–44, or 45–64 years), gender, and recent (past 60 days) mental health treatment, defined as an outpatient visit, inpatient episode, emergency department visit, or other contact with a mental disorder (ICD-9-CM:290–319) diagnosis. We also classified episodes with respect to the presence or absence of a deliberate self-harm diagnosis (E950-E959, “recent self-harm”) during the 60 days before the emergency visit. For the analysis of emergency mental health assessments, we further classified treatment episodes by a diagnosis of depressive (296.2–296.3, 298.0, 300.4, 311), bipolar (296.0, 296.4, 296.5–296.8), anxiety (300.0, 300.2, 300.3, 293.84, 300.83, 309.81), adjustment (308.3, 309.0–309.2, 309.4, 309.9), schizophrenia and related disorders (295, 297–299), substance use (291, 292, 303–305), personality (301), and other mental (290–319, not otherwise classified) disorder during the 60 days before the emergency visit. In the analysis of follow-up outpatient mental health care, mental disorder groups were related to clinical diagnoses given at the emergency department visit.

We classified self-harm events by method into high, low, or unknown lethality.24 High lethality methods included firearm, drowning, suffocation, fall, fire, and motor vehicle, whereas low lethality methods included cutting and poisoning. We classified unspecified or poorly specified codes as “other or unknown.” We also created separate strata for cutting and poisoning.

We used Area Resource File data to characterize the residence of patients by county per capita annual income (low < $25 000; medium = $25 000–40 000; high > $40 000), percentage of county population in poverty (low = 0%–14%; medium = 15%–19%; high ≥ 20%), and county population per square mile (low ≤ 400; high > 400). These county-level variables are significantly related to receiving outpatient mental health care following inpatient admission.25 In contrast to the claims-based patient-level variables, the Area Resource File data are county level.

Analytic Plan

We first compared the characteristics of privately and Medicaid-insured emergency department self-harm episodes with respect to patient characteristics. We then determined the percentages of privately insured and Medicaid-insured patients discharged to the community overall and stratified by patient characteristics and geographic variables. We fit a logistic regression for each characteristic. We first forced each independent variable of interest into each model along with covariates for patient age (21–34 vs 45–64 and 35–44 vs 45–64 years), gender (men vs women), any recent outpatient (present vs absent), inpatient (present vs absent), and emergency department visit (present vs absent) mental health care. We examined interactions between payment status and each independent variable to assess whether the odds ratio (OR) of insurance status related to the outcome significantly differed across strata of the independent variable.

Among patients discharged to the community, we determined stratified percentages of mental health assessment for both study groups across patient characteristics. Using logistic regression models, we estimated associations between each characteristic and the odds of mental health assessment in the emergency department with Medicaid-insured patient episodes as the reference group. We performed similar analyses with outpatient mental health care within 30 days of the emergency department visit as the dependent variable. We have presented results as adjusted ORs (AORs) with associated 95% confidence intervals (CIs). We consider group differences with an OR of 1.30 or more or an OR of 0.70 or less to be potentially substantial from a policy perspective. Because individual patients were permitted to contribute more than 1 episode to the analysis, the observations are nonindependent. Accordingly, we used the SAS 9.2 (SAS Institute, Cary, NC) SURVEYLOGISTIC procedure to accommodate the clustering of observations.

RESULTS

Compared with privately insured patients (n = 2352), Medicaid-insured patients (n = 8226) were younger, more likely to have received recent mental health care in each of the settings, and more likely to have received recent treatment of each of the diagnostic groups, except adjustment disorder. Medicaid-insured patients were also more likely than were privately insured patients to have been recently treated for deliberate self-harm. In relation to self-harm events of privately insured patients, those of Medicaid-insured patients were more likely to involve cutting and less likely to involve poisoning (Table 1).

TABLE 1—

Patient Characteristics: National Medicaid Analytic Extract Files (2006) and MarketScan Research Databases (2005–2007)

Emergency Patients With Deliberate Self-Harm
Characteristic Privately Insured (n = 2352), % Medicaid Insured (n = 8226), % χ2 df P
Age, y 68.3 2 < .001
 21–34 38.8 46.7
 35–44 29.9 30.5
 45–64 31.3 22.8
Gender 0 1 .99
 Men 32.1 32.1
 Women 67.9 67.9
Recent mental health carea
 Any 47.1 59.1 108.5 1 < .001
 Outpatient 42.7 52.0 57.0 1 < .001
 Inpatient 8.9 13.5 33.1 1 < .001
 Emergency 11.5 25.0 178.7 1 < .001
Recent clinical mental disorder diagnosisa
 Depression 30.3 34.0 10.5 1 .001
 Bipolar 11.1 17.8 50.7 1 < .001
 Schizophrenia 3.3 15.6 224.6 1 < .001
 Substance use 11.1 23.1 148.2 1 < .001
 Anxiety 17.6 26.1 64.7 1 < .001
 Adjustment 4.4 3.9 1.4 1 .24
 Personality 2.2 6.5 54.8 1 < .001
 Other 6.7 12.6 54.7 1 < .001
Recent self-harma 11.0 12.0 6.2 1 .01
Current self-harm methodb 5.4 2 .07
 Low lethality 91.3 90.6
 High lethality 3.4 2.9
 Other or unknown 5.3 6.5
Current low-lethality self-harm method
 Cutting 10.2 15.2 32.1 1 < .001
 Poisoning 81.4 75.8 28.8 1 < .001
Open in a new tab
a

On the basis of 60 d before emergency department visit.

b

High-lethality methods include firearm, drowning, suffocation, fall, fire, and motor vehicle; low lethality methods include cutting and poisoning; unknown includes unspecified or poorly specified.

Discharge to the Community

A significantly greater proportion of Medicaid- than privately insured patients were discharged to the community (Table 2). Similar relationships were evident for each age group, for both genders, and without regard to recent mental health treatment. Among patients who had received outpatient mental health care in the past 60 days, 36.4% of privately insured and 60.1% of publicly insured patients were discharged to the community. The association between insurance status and community discharge was significantly stronger for patients with than without recent mental health care use.

TABLE 2—

Rates of Discharge, Total and Stratified by Patient Characteristics: National Medicaid Analytic Extract Files (2006) and MarketScan Research Databases (2005–2007)

Emergency Patients With Deliberate Self-Harm Discharged to Community
Characteristic Privately Insured, % Medicaid Insured, % Privately Insured Patients, No. Medicaid-Insured Patients, No. AOR (95% CI)
Total 46.9 62.7 2352 8226 0.52* (0.47, 0.58)
Age, y
 21–34 52.6 67.1 912 3840 0.52* (0.44, 0.61)
 35–44 45.3 60.7 704 2513 0.50* (0.42, 0.60)
 45–64 41.3 56.3 736 1873 0.54* (0.45, 0.65)
Gender
 Men 49.1 62.0 755 2639 0.58* (0.48, 0.69)
 Women 45.8 63.0 1597 5587 0.49* (0.44, 0.56)
Recent mental health carea
 None 55.1 66.4 1245 3361 0.66* (0.67, 0.75)
 Any 37.7 60.1 1107 4865 0.41* (0.36, 0.48)
 Outpatient 36.4 60.1 1004 4280 0.39* (0.34, 0.46)
 Inpatient 30.6 50.2 209 1109 0.43* (0.31, 0.60)
 Emergency 44.4 61.0 270 2053 0.48* (0.36, 0.63)
Recent clinical mental disorder diagnosisab
 Depressiond 37.2 57.9 712 2800 0.44* (0.37, 0.53)
 Bipolard 29.4 60.2 262 1462 0.30* (0.22, 0.41)
 Anxietyd 37.6 62.1 415 2144 0.39* (0.30, 0.49)
 Adjustment 42.3 57.4 104 317 0.52* (0.30, 0.87)
 Schizophrenia 39.7 59.4 78 1286 0.51* (0.31, 0.83)
 Substance use 40.8 58.8 262 1904 0.51* (0.38, 0.67)
 Personality 35.3 59.9 51 538 0.57* (0.28, 1.16)
 Otherd 35.4 65.0 158 1034 0.30* (0.20, 0.45)
Recent self-harmad
 Yes 45.0 64.5 258 1061 0.45* (0.33, 0.60)
 No 47.1 62.4 2094 7165 0.53* (0.33, 0.60)
Current self-harm methodc
 Low lethality 44.9 62.3 2148 7454 0.49* (0.44, 0.55)
 High lethality 58.2 57.4 79 235 1.15* (0.67, 1.99)
 Other or unknown 73.6 70.0 125 537 1.03* (0.62, 1.73)
Current low-lethality self-harm method
 Cutting 51.7 71.4 240 1251 0.42* (0.31, 0.57)
 Poisoning 44.0 60.4 1916 6235 0.51* (0.46, 0.57)
Open in a new tab

Note. AOR = adjusted odds ratio; CI = confidence interval. Regressions control for patient age, patient gender, any outpatient mental health care in past 60 d, any inpatient mental health care in past 60 d, and any emergency department care in past 60 d. Separate models for each strata: insurance (Medicaid vs private) is the independent variable of interest, and Medicaid is the reference group.

a

On the basis of 60 d before emergency department visit.

b

For mental disorder variables, reference group is treatment episodes without the disorder.

c

High-lethality methods include firearm, drowning, suffocation, fall, fire, and motor vehicle; low lethality methods include cutting and poisoning; unknown includes unspecified or poorly specified.

d

Indicates significant interaction (P < .05) with insurance status.

*P < .05.

For each clinical diagnostic group examined, community discharge was also significantly more common among Medicaid- than among privately insured patients. This relationship was especially evident for patients who had been recently treated for bipolar disorder, as indicated by the significant interaction between payment source, bipolar diagnosis, and community discharge. Similar interactions were evident between payment source and recent treatment of depression, anxiety, and other mental disorders (Table 2).

Mental Health Assessment

A significantly larger percentage of discharged patients with self-harm who were privately rather than Medicaid insured received an emergency mental health assessment (Table 3). In the adjusted models, the odds of emergency mental health assessment was significantly greater for privately than for Medicaid-insured discharged patients who had been recently treated for schizophrenia, depressive disorders, and anxiety disorders as well as those whose current self-harm method was ingestion. The association between payment source and emergency mental health assessment was significantly stronger for patients without than with recent inpatient mental health treatment and for patients without than with a highly lethal self-harm method (Table 3).

TABLE 3—

Rates of Mental Disorder Assessment Discharged to the Community, Total and Stratified by Patient Characteristics: National Medicaid Analytic Extract Files (2006) and MarketScan Research Databases (2005–2007)

Mental Disorder Assessment in Emergency Department
Characteristic Privately Insured, % Medicaid Insured, % Privately Insured Patients, No. Medicaid-Insured Patients, No. AOR (95% CI)
Total 57.3 47.8 1103 5155 1.58* (1.38, 1.81)
Age, y
 21–34 57.9 47.0 480 2575 1.68* (1.37, 2.06)
 35–44 58.9 47.8 319 1525 1.70* (1.33, 2.18)
 45–64 54.6 49.9 304 1055 1.32* (1.01, 1.72)
Gender
 Men 52.8 47.5 371 1636 1.36* (1.07, 1.72)
 Women 59.6 48.0 732 3519 1.69* (1.42, 2.01)
Recent mental health carea
 None 56.0 43.0 686 2233 1.70* (1.43, 2.02)
 Any 59.5 51.5 417 2922 1.43* (1.16, 1.77)
 Outpatient 59.7 51.1 365 2572 1.47* (1.18, 1.84)
 Inpatientd 51.6 56.9 64 557 0.93* (0.54, 1.60)
 Emergency 62.5 55.4 120 1252 1.37* (0.92, 2.04)
Recent mental disorder diagnosisab
 Depression 61.9 51.7 265 1620 1.62* (1.23, 2.12)
 Bipolar 54.5 51.9 77 880 1.16* (0.72, 1.89)
 Anxiety 59.6 51.6 156 1331 1.47* (1.04, 2.07)
 Adjustment 56.8 48.9 44 182 1.46* (0.71, 3.01)
 Schizophrenia 77.4 49.3 31 764 3.72* (1.63, 8.48)
 Substance use 63.6 54.5 107 1119 1.51* (0.99, 2.29)
 Personality 61.1 53.4 18 322 1.43* (0.45, 4.61)
 Other 57.1 52.5 56 672 1.24* (0.72, 2.15)
Recent self-harma
 Yes 64.7 49.4 116 684 2.00* (1.29, 3.11)
 No 56.4 47.6 987 4471 1.54* (1.33, 1.77)
Current self-harm methodc
 Low lethality 59.1 47.8 965 4644 1.67* (1.45, 1.93)
 High lethalityd 37.0 42.2 46 46 0.93* (0.42, 2.04)
 Other or unknown 48.9 50.3 92 376 1.38* (0.83, 2.30)
Current low-lethality self-harm method
 Cutting 51.6 46.8 124 893 1.27* (0.87, 1.87)
 Poisoning 60.3 48.0 843 3767 1.74* (1.49, 2.03)
Open in a new tab

Note. AOR = adjusted odds ratio; CI = confidence interval. Regressions control for patient age, patient gender, any outpatient mental health care in past 60 d, any inpatient mental health care in past 60 d, and any emergency department care in past 60 d. Separate models for each strata: insurance (Medicaid vs private) is the independent variable of interest, and Medicaid is the reference group.

a

On the basis of 60 d before emergency department visit.

b

For mental disorder variables, reference group is treatment episodes without the disorder.

c

High-lethality methods include firearm, drowning, suffocation, fall, fire, and motor vehicle; low lethality methods include cutting and poisoning; unknown includes unspecified or poorly specified.

d

Significant interaction (P < .05) with insurance status.

*P < .05.

Outpatient Mental Health Care

Privately insured patients (41.2%) were significantly less likely than were Medicaid-insured patients (52.9%) to receive outpatient mental health care in the 30 days following community discharge (Table 4). The odds of follow-up care were strongly related to prior mental health care use. In both payment groups, relatively low rates of follow-up were evident for patients without recent mental health service use. Payment source had a significantly stronger association with emergency mental health assessment among patients with than without recent mental health service use (Table 4).

TABLE 4—

Rates of Outpatient Mental Health Care, Total and Stratified by Patient Characteristics: National Medicaid Analytic Extract Files (2006) and MarketScan Research Databases (2005–2007)

Mental Health Follow-Up of Outpatient
Characteristic Privately Insured, % Medicaid Insured, % Privately Insured Patients, No. Medicaid-Insured Patients, No. AOR (95% CI)
Total 41.2 52.9 1103 5155 0.78* (0.67, 0.92)
Age, y
 21–34 37.7 50.3 480 2575 0.75* (0.59, 0.96)
 35–44 44.5 55.1 319 1525 0.88* (0.65, 1.18)
 45–64 43.1 56.1 304 1055 0.73* (0.54, 0.99)
Gender
 Men 34.8 50.2 371 1636 0.66* (0.50, 0.87)
 Women 44.4 54.1 732 3519 0.85* (0.70, 1.03)
Recent mental health care before emergency department visitac
 None 29.7 31.7 686 2233 0.90* (0.74, 1.10)
 Any 60.0 69.1 417 2922 0.63* (0.49, 0.80)
 Outpatientc 64.7 74.3 365 2572 0.61* (0.47, 0.79)
 Inpatientc 57.8 67.1 64 557 0.56* (0.31, 1.00)
 Emergency 44.2 63.1 120 1252 0.54* (0.34, 0.86)
Mental disorder diagnosis in emergency department
 Depression 48.4 59.2 384 1041 0.80* (0.61, 1.06)
 Bipolar 61.8 69.1 68 269 0.95* (0.51, 1.77)
 Anxiety 45.6 58.9 171 817 0.71* (0.49, 1.04)
 Adjustment 45.7 35.1 46 97 1.56* (0.68, 3.60)
 Schizophrenia 58.8 67.6 17 188 0.87* (0.29, 2.61)
 Substance use 44.2 51.4 224 888 0.98* (0.70, 1.35)
 Personality 53.8 67.9 13 81 1.09* (0.29, 4.20)
 Other 48.8 58.5 41 130 1.10* (0.46, 2.62)
Emergency mental health assessment
 No emergency mental health assessmentc 32.9 49.8 471 2690 0.59* (0.46, 0.75)
 No recent mental health care 21.5 28.5 302 1273 0.68* (0.50, 0.92)
 Recent mental health care 53.3 68.9 169 1417 0.49* (0.34, 0.70)
 Emergency mental health assessment 47.3 56.2 632 2465 0.90* (0.74, 1.08)
 No recent mental health care 36.2 35.8 384 960 1.01* (0.78, 1.29)
 Recent mental health care 64.5 69.3 248 1505 0.75* (0.55, 1.03)
Recent self-harma
 Recent self-harm 50.0 54.1 116 684 0.76* (0.46, 1.24)
 No recent self-harm 40.8 52.7 987 4471 0.78* (0.67, 0.91)
Current self-harm methodb
 Low lethality 44.4 53.6 965 4644 0.84* (0.70, 0.99)
 High lethality 26.1 42.2 46 135 0.62* (0.28, 1.38)
 Other or unknownc 15.2 47.6 92 376 0.34* (0.17, 0.67)
Current low-lethality self-harm method
 Cutting 40.3 56.1 124 893 0.76* (0.47, 1.23)
 Poisoning 45.0 53.0 843 3767 0.85* (0.71, 1.02)
Open in a new tab

Note. AOR = adjusted odds ratio; CI = confidence interval. Regressions control for patient age, patient gender, any outpatient mental health care in past 60 d, any inpatient mental health care in past 60 d, and any emergency department care in the past 60 d. Medicaid is the reference group.

a

On the basis of 60 d before emergency department visit.

b

High-lethality methods include firearm, drowning, suffocation, fall, fire, and motor vehicle; low lethality methods include cutting and poisoning; unknown includes unspecified or poorly specified.

c

Indicates significant interaction (P < .05) with insurance status.

*P < .05.

A significant interaction existed between payment source and emergency mental health assessment with mental health follow-up care. When no emergency mental health assessment was performed, privately insured patients were significantly less likely than were Medicaid-insured patients to have follow-up.

Privately (41.8%) and Medicaid- (44.5%) insured patients in high poverty counties had similar rates of emergency mental health assessment (AOR = 0.98; 95% CI = 0.58, 1.65). For other county-level strata examined including low and medium county population in poverty, the 3 population density strata, and the 3 income strata, the AORs of emergency mental health assessments were significantly greater for privately than for Medicaid-insured patients (data not shown).

DISCUSSION

Privately insured patients with self-harm were more likely than were comparable Medicaid-insured patients to be admitted to the hospital. Among those who were discharged, privately insured patients were also more likely than were Medicaid-insured patients to have received a mental health assessment in the emergency department, but they were less likely to follow up with outpatient mental health care. These findings suggest that health insurance influences the delivery of acute mental health care of deliberate self-harm.

The decision to admit or discharge deliberate patients with self-harm from the emergency department is known to be related to suicide risk. Several traditional suicide risk factors—including male gender,26,27 older patient age,28 previous psychiatric hospital admission,26,29 current psychiatric treatment,27,28 psychosis,24,29 use of a lethal self-harm method,26,29 and previous deliberate self-harm events26,27,29—have been demonstrated to influence disposition decisions. The possibility that source of payment also contributes to the emergency department disposition raises a less understood and potentially important aspect of service delivery.

For a wide range of patients with self-harm, privately insured patients were significantly more likely to be admitted to the hospital than were their Medicaid-insured counterparts. Some of the largest payment group differences in disposition were among patients at high risk of suicide. Although depressive disorders are a powerful suicide risk factor,30 insurance status was significantly more strongly associated with emergency department disposition for patients with than without recent depression treatment. We observed similar interactions of payment source with treatment of bipolar disorder,31 recent self-harm,9,28 and any outpatient mental health care.32 Payment group disposition differences in high-risk patient groups underscore the importance of determining and addressing clinical and service factors that drive this practice variation.

Privately insured discharged patients were also more likely than were Medicaid-insured patients to receive a mental health assessment in the emergency department. This payment group difference extended to patients with recent prior self-harm,33 depression,34 and schizophrenia,35 who are at high risk for repeated deliberate self-harm and suicide. These findings should stimulate programmatic reforms to improve access to emergency mental health services, especially in hospitals that serve substantial numbers of Medicaid-insured patients.

Lethality of self-harm method positively correlates with suicidal intent36 and future suicide risk.26,29 On this basis, it might be expected that mental health assessments would be routine in the emergency evaluation of self-injuries from firearms or other highly lethal methods. Yet only a minority of discharged patients who used highly lethal self-harm methods received emergency mental health assessments. Although the reasons for this low level of attention to the mental health status of these patients cannot be determined from a claims-based analysis, it is possible that the acute physical health threats posed by these presumably more serious self-injuries diverts clinical attention of some emergency department staff away from mental health aspects of care.

Lower rates of hospital admission and mental health assessment of Medicaid- than privately insured patients with self-harm exist within a broader context of disparities by payment source in access and quality of care.1,3,4 The differences in processes of care across sources of payments may exist either between or within individual hospitals. Research from the United Kingdom37 and Finland38 indicates that individual hospitals vary in the proportion of patients with self-harm who receive psychosocial assessments and hospital admission. Medicaid patients with self-harm may receive a disproportionate share of their care from hospitals with limited access to inpatient beds and therefore have a higher clinical threshold for hospital admission. Such hospitals may also have few if any emergency mental health services. Detailed hospital-level studies are needed to isolate the contribution of between- and within-hospital variation to payment source variation in acute self-harm management.

During the 30 days following discharge, 41.2% of privately insured and 52.9% of Medicaid-insured patients received outpatient mental health care. A Finnish study reported that 54.0% of discharged patients with self-harm from Helsinki hospitals received outpatient care within 30 days.38 A Danish study reported that 57.0% of emergency department patients with self-harm from Copenhagen completed referrals to a suicide prevention clinic.39 Both of these countries have universal public health care systems and national recommendations to evaluate and offer treatment to all individuals who attempt suicide.

Contrary to our expectations, a greater proportion of Medicaid- than privately insured patients received timely outpatient mental health care. Previous research indicates that community clinics more readily offer postemergency department follow-up mental health appointments to privately than to Medicaid-insured patients.40 Yet Medicaid-insured patients may have access to more extensive behavioral health services than do privately insured patients.41 For example, Medicaid, but not private plans, typically cover case management, crisis management, and family support and peer support mental health services.41 As a result of limited access to inpatient care for Medicaid-insured patients, emergency staff may also be especially motivated to help Medicaid-insured patients connect with outpatient mental health care.

Health care reform offers opportunities to improve the integration of emergency and outpatient care. Private plans have begun experimenting with new models to improve transitions between hospital and outpatient services.42 In addition, the Affordable Care Act43 substantially increases reimbursement to state Medicaid programs for specialized health home programs for Medicaid beneficiaries with serious mental disorders and other chronic conditions. Health homes must provide care coordination and transitional care services.44 It is important to track the effects of health care reform on links between emergency and outpatient mental health care.

Prior mental health service appears to play an important role in shaping follow-up mental health care.17 Patients with self-harm who had not received recent mental health care had a particularly low rate of follow-up mental health care. Significant payment group differences in follow-up care were evident only among patients who had received recent mental health care. Among these patients, a significant group difference in follow-up care was apparent only among patients who had not received an emergency mental health assessment. Recent mental health service use, therefore, appears to be more important than does emergency department mental health assessments in accounting for the greater likelihood of patients with Medicaid than those with private insurance to receive follow-up mental health care.

Limitations

This study has several limitations. First, data were not available concerning several clinically important factors such as the patient's intent to engage in repeated deliberate self-harm, intent to complete suicide,29 willingness to receive mental health care, and level of social support. These and other unmeasured factors may account for the observed differences across payment groups in the likelihood of inpatient admission, emergency department assessment, and mental health care follow-up. Second, concerns exist over the validity and completeness of e-codes as a measure of deliberate self-harm. The quality of e-coding may vary across states45 and across private and public health plans. In addition, the codes do not differentiate deliberate self-harm of suicidal from nonsuicidal intent, though small studies support high positive predictive value with respect to medical record notations46 and expert assessment47 of attempted suicide. Third, diagnoses were derived from clinician judgment without independent expert validation. Fourth, data were not available on the emergency mental health assessment and diagnoses of patients who were admitted for inpatient treatment. Fifth, the analyses do not capture services that are provided free, not reimbursed, or paid for out of pocket. Finally, an insurance claim for a mental disorder or a mental health assessment procedure does not ensure that a thorough mental health assessment was performed. However, an absence of these services suggests a missed opportunity to provide mental health care.

Conclusions

Deliberate self-harm poses a high risk for future deliberate self-harm and suicide.8–10 Following medically injurious deliberate self-harm, Medicaid-insured patients are substantially less likely than are privately insured patients to be admitted for inpatient treatment and, among those who are discharged to the community, Medicaid-insured patients are also less likely to have received an emergency mental health assessment. Efforts are needed to expand the availability and access of emergency mental health services to all patients with self-harm. As implementation of the Affordable Care Act moves forward, continuity in the acute care management of deliberate self-harm should be a key target for quality of care improvement.

Roughly one half of patients who are discharged from emergency departments following an episode of deliberate self-harm did not receive timely follow-up outpatient mental health care. Problems with transitions to outpatient mental health care were especially evident for privately insured patients and patients who were not receiving mental health care before the self-harm event. Successful models of integrating care in other areas of medicine47 may inform efforts to improve the transition from emergency to outpatient mental health care. Alongside systematic efforts to coordinate care, simple clinical interventions, including provision of specific appointments and telephone reminders, may improve adherence to outpatient referrals. A robust emergency mental health research agenda is needed to promote the development and testing of feasible interventions to improve mental health assessment, treatment, and referral coordination. Although a few model programs have demonstrated encouraging results in improving linkage with outpatient care,49,50 special efforts will likely be needed to extend effective models to the large number of privately and publicly insured patients who go to emergency departments following a deliberate self-harm event.

Acknowledgments

The American Foundation for Suicide Prevention of Columbia University supported this study (grant no. 2DIG-00005-1207-1208).

Human Participant Protection

The institutional review board of New York State Psychiatric Institute determined that this study did not require approval because it did not meet the federal definition of human subjects research.

References

  • 1.Amundson GM, O'Connor PJ, Solberg LIet al. Diabetes care quality: insurance, health plan, and physician group contributions. Am J Manag Care. 2009;15(9):585–592. [PubMed] [Google Scholar]
  • 2.Calvin JE, Rose MT, Chen AYet al. Insurance coverage and care of patients with non-ST-segment elevation acute coronary syndromes. Ann Intern Med. 2006;145(10):739–748. [DOI] [PubMed] [Google Scholar]
  • 3.Landon BE, Schneider EC, Normand SL, Scholle SH, Pawlson LG, Epstein AM. Quality of care in Medicaid managed care and commercial health plans. JAMA. 2007;298(14):1674–1681. [DOI] [PubMed] [Google Scholar]
  • 4.Thompson JW, Ryan KW, Pinidiya SD, Bost JE. Quality of care for children in commercial and Medicaid managed care. JAMA. 2003;290(11):1486–1493. [DOI] [PubMed] [Google Scholar]
  • 5.Mann JJ, Apter A, Bertolote Jet al. Suicide prevention strategies: a systematic review. JAMA. 2005;294(16):2064–2074. [DOI] [PubMed] [Google Scholar]
  • 6.Brådvik L, Mattisson C, Bogren M, Nettelbladt P. Long-term suicide risk in depression in the Lundby cohort 1947–1997—severity and gender. Acta Psychiatr Scand. 2008;117(3):185–191. [DOI] [PubMed] [Google Scholar]
  • 7.Goldstein RB, Black DW, Nasrallah A, Winokur G. The prediction of suicide. Sensitivity, specificity, and predictive value of a multivariate model applied to suicide among 1906 patients with affective disorders. Arch Gen Psychiatry. 1991;48(5):418–422. [DOI] [PubMed] [Google Scholar]
  • 8.Hawton K, Zahl D, Weatherall R. Suicide following deliberate self-harm: long-term follow-up of patients who presented to a general hospital. Br J Psychiatry. 2003;182:537–542. [DOI] [PubMed] [Google Scholar]
  • 9.Cooper J, Kapur N, Webb Ret al. Suicide after deliberate self-harm: a 4-year cohort study. Am J Psychiatry. 2005;162(2):297–303. [DOI] [PubMed] [Google Scholar]
  • 10.Chen VCH, Tan KLT, Chung-Ying Cet al. Mortality and suicide after self-harm: community cohort study in Taiwan. Br J Psychiatry. 2011;198(1):31–36. [DOI] [PubMed] [Google Scholar]
  • 11.Nock MK, Hwang I, Sampson NA, Kessler RC. Mental disorders, comorbidity and suicidal behavior: results from the National Comorbidity Survey Replication. Mol Psychiatry. 2010;15(8):868–876. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Beautrais AL, Joyce PR, Mulder RT, Fergusson DM, Deavoll BJ, Nightingale SK. Prevalence and comorbidity of mental disorders in persons making serious suicide attempts: case-control study. Am J Psychiatry. 1996;153(8):1009–1014. [DOI] [PubMed] [Google Scholar]
  • 13.National Institute for Clinical Excellence. Self-Harm: The Short Term Physical and Psychological Management and Secondary Prevention of Self-Harm in Primary and Secondary Care. London: National Institute for Clinical Excellence; 2004. [Google Scholar]
  • 14.American Psychiatric Association. Practice Guideline for the Assessment and Treatment of Patients With Suicidal Behaviors. Arlington, VA; 2003. [PubMed] [Google Scholar]
  • 15.Baraff LJ, Janowicz N, Asarnow JR. Survey of California emergency departments about practices for management of suicidal patients and resources available for their care. Ann Emerg Med. 2006;48(4):452–458. [DOI] [PubMed] [Google Scholar]
  • 16.Olfson M, Marcus SC, Bridge JA. Emergency treatment of deliberate self-harm. Arch Gen Psychiatry. 2012;69(1):80–88. [DOI] [PubMed] [Google Scholar]
  • 17.Adamson DM, Chang S, Hansen LG. Health Research Data for the Real World: The MarketScan Databases (White Paper). Ann Arbor, MI: Thomson Healthcare; July 2006.
  • 18.DeNavas-Walt C, Proctor BD, Smith JC. Income, Poverty, and Health Insurance Coverage in the United States: 2010 (US Census Bureau, Current Population Reports). Washington, DC: US Government Printing Office; 2011. Series P60-239.
  • 19.Klees BS, Wolfe CJ, Curtis CA; Office of the Actuary, Centers for Medicare and Medicaid Services. Brief Summaries of Medicare & Medicaid, Title XVIII and Title XIX of the Social Security Act. Available at: https://www.cms.gov/medicareprogramratesstats/downloads/MedicareMedicaidSummaries2010.pdf. Accessed October 20, 2011.
  • 20.Claxton G, DiJulio B, Whitmore Het al. Health benefits in 2010: premiums rise modestly, workers pay more toward coverage. Health Aff. 2010;29(10):1942–1950. [DOI] [PubMed] [Google Scholar]
  • 21.Hsia RY, MacIsaac D, Baker LC. Decreasing reimbursements for outpatient emergency department visits across payer groups from 1996 to 2004. Ann Emerg Med. 2008;51(3):265–274. [DOI] [PubMed] [Google Scholar]
  • 22.National Center for Health Workforce Analysis. Area Resource File. 2006. [Google Scholar]
  • 23.National Committee for Quality Assurance. The State of Health Care Quality. Washington, DC; 2008. [Google Scholar]
  • 24.Runeson B, Tidemalm D, Dahlin M, Lictenstein P, Långström N. Method of attempted suicide as predictor of subsequent successful suicide: national long term cohort study. BMJ. 2010;341:c3222. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Olfson M, Marcus SC, Doshi JA. Continuity of care after inpatient discharge of schizophrenia patients. J Clin Psychiatry. 2010;71(7):831–838. [DOI] [PubMed] [Google Scholar]
  • 26.Miret M, Nuevo R, Morant Cet al. The role of suicide risk in the decision for psychiatric hospitalization after a suicide attempt. Crisis. 2011;32(2):65–73. [DOI] [PubMed] [Google Scholar]
  • 27.Goldberg JF, Ernst CL, Bird S. Predicting hospitalization versus discharge of suicidal patients presenting to a psychiatric emergency service. Psychiatr Serv. 2007;58(4):561–565. [DOI] [PubMed] [Google Scholar]
  • 28.Suominen K, Isometsä E, Suokas J, Haukka J, Achte K, Lönnqvist J. Completed suicide after a suicide attempt: a 37-year follow-up study. Am J Psychiatry. 2004;161(3):562–563. [DOI] [PubMed] [Google Scholar]
  • 29.Baca-Garcia E, Diaz-Sastre C, Resa EGet al. Variables associated with hospitalization decisions by emergency psychiatrists after a patient's suicide attempt. Psychol Serv. 2004;55(7):792–797. [DOI] [PubMed] [Google Scholar]
  • 30.Hunt IM, Kapur N, Robinson Jet al. Suicide within 12 months of mental health service contact in different age and diagnostic groups: national clinical survey. Br J Psychiatry. 2006;188(2):135–142. [DOI] [PubMed] [Google Scholar]
  • 31.Tondo L, Isacsson G, Baldessarini R. Suicidal behavior in bipolar disorder: risk and prevention. CNS Drugs. 2003;17(7):491–511. [DOI] [PubMed] [Google Scholar]
  • 32.Becker M, Brown L, Ochshorn E, Diamond R. Risk for suicide among Medicaid beneficiaries. Suicide Life Threat Behav. 2009;39(2):172–181. [DOI] [PubMed] [Google Scholar]
  • 33.Kapur N, House A, Dodgson K, May C, Creed F. Effect of general hospital management on repeat episodes of deliberate self poisoning: cohort study. BMJ. 2002;325(7369):866–867. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Sakinofsky I, Roberts RS. Why parasuicides repeat despite problem resolution. Br J Psychiatry. 1990;156:399–405. [DOI] [PubMed] [Google Scholar]
  • 35.Vajda J, Steinbeck K. Factors associated with repeat suicide attempts among adolescents. Aust N Z J Psychiatry. 2000;34(3):437–445. [DOI] [PubMed] [Google Scholar]
  • 36.Harriss L, Hawton K. Suicidal intent in deliberate self-harm and the risk of suicide: the predictive power of the Suicide Intent Scale. J Affect Disord. 2005;86(2–3):225–233. [DOI] [PubMed] [Google Scholar]
  • 37.Bennewith O, Peters TJ, Hawton K, House A, Gunnell D. Factors associated with the non-assessment of self-harm patients attending an accident and emergency department: results of a national study. J Affect Disord. 2005;89(1–3):91–97. [DOI] [PubMed] [Google Scholar]
  • 38.Suominen K, Isometsa E, Martunnen M, Ostamo A, Lönnqvist J. Health care contacts before and after attempted suicide among adolescent and young adult versus older suicide attempters. Psychol Med. 2004;34(2):313–321. [DOI] [PubMed] [Google Scholar]
  • 39.Nordentoft M, Søgaard M. Registration, psychiatric evaluation, and adherence to psychiatric treatment after suicide attempt. Nord J Psychiatry. 2005;59(3):213–216. [DOI] [PubMed] [Google Scholar]
  • 40.Rhodes KV, Vieth TL, Kushner H, Levy H, Asplin BR. Referral without access: for psychiatric services, wait for the beep. Ann Emerg Med. 2009;54(2):272–278. [DOI] [PubMed] [Google Scholar]
  • 41.Garfield RL, Lave JR, Donohue JM. Health reform and the scope of benefits for mental health and substance use disorder services. Psychiatr Serv. 2010;61(11):1081–1086. [DOI] [PubMed] [Google Scholar]
  • 42.AHIP Center for Policy and Research. Innovations in Reducing Preventable Hospital Admissions, Readmissions, and Emergency Room Use; June 2010. Available at: http://www.ahipresearch.org/pdfs/innovations2010.pdf. Accessed October 12, 2011.
  • 43.The Patient Protection and Affordable Care Act. Pub. L. No. 111–148, 124 Stat. 119, H.R. 3590.
  • 44.Henry J. Kaiser Family Foundation. Focus on Health Reform: Medicaid's New “Health Home” Option; January 2011. Pub no. 8316. Available at: http://www.kff.org/medicaid/upload/8136.pdf. Accessed October 12, 2011.
  • 45.Barrett M, Steiner C, Coben J. Healthcare Cost and Utilization Project (HCUP) E Code Evaluation Report. US Agency for Healthcare Research and Quality. HCUP Methods Series Report no. 2004-06; 2004. Available at: http://www.hcup-us.ahrq.gov/ reports/methods.jsp. Accessed October 14, 2011.
  • 46.Simon GE, Savarino J. Suicide attempts among patients starting depression treatment with medications or psychotherapy. Am J Psychiatry. 2007;164(7):1029–1034. [DOI] [PubMed] [Google Scholar]
  • 47.Iribarren C, Sidney S, Jacobs DR, Jr, Weisner C. Hospitalization for suicide attempt and completed suicide: epidemiological features in a managed care population. Soc Psychiatry Psychiatr Epidemiol. 2000;35(7):288–296. [DOI] [PubMed] [Google Scholar]
  • 48.Naylor MD, Feldman PH, Keating Set al. Translating research into practice: transitional care for older adults. J Eval Clin Pract. 2009;15(6):1164–1170. [DOI] [PubMed] [Google Scholar]
  • 49.Chan TC, Killeen JP, Castillo EMet al. Impact of an internet-based emergency department appointment system to access primary care at safety net community clinics. Ann Emerg Med. 2009;54(2):279–284. [DOI] [PubMed] [Google Scholar]
  • 50.McCusker J, Dendukuri N, Tousignant Pet al. Rapid two-stage emergency department intervention for seniors: impact on continuity of care. Acad Emerg Med. 2003;10(3):233–243. [DOI] [PubMed] [Google Scholar]

Articles from American Journal of Public Health are provided here courtesy of American Public Health Association

RESOURCES