ED Supporting Statement_B_091013

Evaluation of Emergency Department Crisis Center Follow-up—New

1. Statistical Methods

B1. Respondent Universe and Sampling Methods

There are 159 crisis centers in the National Suicide Prevention Lifeline Network. The proposed data collection will be contained at two participating crisis centers in collaboration with two emergency departments. Two crisis centers conducting follow-up with suicidal callers and emergency department patients will participate in the data collection. The universe of data will be identified at the time of each abstraction. If the total potential sample size exceeds the desired sample size for the pilot, the emergency departments will be asked to randomly select the sample. If the universe is less than the desired sample size, emergency departments will be asked to send 100% of the cases to the evaluation team. Individuals in the collaboration sample who received crisis center follow-up will be compared with matched individuals from the pre-collaboration sample, and assessed for emergency department readmissions within one year of their initial admission. Whenever possible, individuals will be matched on gender, age, ethnicity, diagnosis code (indicating means of attempt), prior suicide attempts, and discharge disposition (or as many of these data elements as each emergency department is able to provide). Priority will be given to matching on gender and age. Our sample size was determined based on what participating emergency departments and crisis centers indicated would be feasible, as well as on calculations using Fleiss (1981) for a simple analysis of the difference between proportions. Subgroup analyses using demographic and lethality variables will be performed on an exploratory basis, to the extent permitted by the data available to us.

B2. Information Collection Procedures

Trained hospital staff will review patient data to identify appropriate patient data for abstraction. Data from 2,000 patient records will be abstracted. Approximately 1,000 patient data records will be abstracted for the two years prior to collaboration between the participating emergency department and crisis center and 1,000 for two years following collaboration between the two organizations. Data will be abstracted once for the pre-collaboration group and once for the collaboration group. Data will be abstracted from relevant electronic medical records by two hospital staff through the Hospital Data Abstraction Form. Demographic and historical data (e.g., patient ID, date of admission¹, gender, age, diagnosis code, prior suicide attempts) will be abstracted along with prior and subsequent emergency department admission data (e.g., date of admission, diagnosis code, discharge status). In addition, 2 crisis center staff, one from each participating crisis center, will abstract companion patient data through the Crisis Center Data Abstraction Form. Information will include the patient ID, the name of the referring hospital, the date of the referral, and whether clinical contact was made with the patient. This data will be abstracted once, at the end of the collaboration period.

For each patient who is admitted to one of the participating emergency departments for suicidal behavior during the pre-collaboration or collaboration period, data on emergency department readmissions of the same patient during the 365 days following the patient’s initial admission will be obtained. Crisis center follow-up data (i.e., whether or not a patient receives follow-up, and date(s) of follow-up contact, if any) will only be obtained for collaboration period samples. All of the analyses will rely on data that is already collected by the emergency departments and crisis centers in the course of their clinical interventions. As such, the timing of the three data abstractions, whereby pre-existing data is transmitted from the emergency departments and crisis centers to evaluation staff, is based on convenience, and does not impact the study design.

This evaluation is subcontracted via ICF Macro to evaluators at Columbia University. All project data will be held by and in the possession and control of these evaluators. Protocols designed to protect patient privacy and the confidentiality of data collected for evaluation purposes have been reviewed by the Institutional Review Board (IRB) of the Columbia University Department of Psychiatry and the New York State Psychiatric Institute. IRB approval was granted on April 29, 2013. The IRB considers the data elements requested by this evaluation to constitute a “limited data set” under HIPAA, and has granted the project waivers of HIPAA authorization and consent.

SAMHSA is likewise requesting a waiver of consent and a waiver of HIPAA authorization to allow Columbia University evaluators to collect and analyze limited data extracted from hospital and crisis center records. SAMHSA is seeking the approval of the waiver of consent, based on Federal Regulations (Title 45; Part 46, Article 46.116(d): (1) this research involves no more than minimal risk (i.e., no information is obtained that would not have otherwise been obtained during the routine course of the clinical crisis intervention); (2) the waiver will not adversely affect the rights and welfare of the subjects (i.e., with the exception of the hospital name, crisis center name, and date(s) of ED admission, all data will be de-identified before it is provided to us by the hospitals and the crisis centers. It will not be possible for us to link the data to further identifiable information. The hospitals and crisis centers have links to subject identifiers, but SAMHSA will not have access to these links. Thus, the risk of the loss of confidentiality is minimal); and (3) the research could not practicably be carried out without the waiver (i.e., there is no other way to get this information).

Central to these waiver requests are the facts that evaluation personnel will have no direct contact with patients, no way of contacting them, and no way of identifying who they are. Again, patients’ contact information and other identifying information used by the emergency departments and crisis centers for clinical purposes will not be shared with evaluation staff. Evaluation staff will not have access to any elements of the patients’ medical records other than the specific elements listed on our two abstraction forms.

Table 3 summarizes the information collection procedures across all components of the evaluation.

TABLE 3

Procedures for the Collection of Information

Measure	Indicators	Data Source(s)	Method	When Collected
Hospital Data Abstraction Form	Patient ID Demographic information Historical data Discharge status Prior suicide attempts Prior emergency department admissions for suicidal behavior Subsequent emergency department admissions for suicidal behavior Patient acceptance of crisis center referral	Hospital data	Review of existing data	Once for two year pre collaboration period and once for two year collaboration period
Crisis Center Data Abstraction form	Patient ID Date of referral Clinical contact (Y/N) Date(s) of clinical contact, if any	Crisis center data	Review of existing data	Once for two year collaboration period

B3. Methods to Maximize Response Rates

The directors of crisis centers and collaborating hospitals have agreed to participate and will secure hospital institutional review board (IRB) approval before evaluation activities commence. There are no direct respondents associated with this data collection effort.

B4. Tests of Procedures

The Hospital Data Abstraction Form and the Crisis Center Data Abstraction Form were developed by Columbia University consultants. All abstraction form measures have been reviewed by experts in the field of mental health and piloted to determine burden levels.

B5. Statistical Consultants

The evaluator has full responsibility for the development of the overall statistical design and assumes oversight responsibility for data collection and analysis for the evaluation. Training and monitoring of data collection will be provided by the evaluator. The following individuals are primarily responsible for overseeing data collection and analysis:

Madelyn S. Gould, Ph.D., M.P.H.

Professor,

Psychiatry and Public Health (Epidemiology)

Columbia University/NYSPI

1051 Riverside Drive, Unit 72

New York, NY 10032

212-543-5329

Alison Lake, M.A.

Psychiatry and Public Health (Epidemiology)

Columbia University/NYSPI

1051 Riverside Drive, Unit 72

New York, NY 10032

212-543-6714

The SAMHSA project officer responsible for receiving and approving deliverables is:

James Wright, MS, LCPC

Suicide Prevention Branch

Center for Mental Health Services

SAMHSA

1 Choke Cherry Road

Rockville, MD 20857

240-276-1854

James.wright@samhsa.hhs.gov

References

Andrews, G., & Sunderland, M. (2009). Telephone case management reduces both distress and psychiatric hospitalization. Australian and New Zealand Journal of Psychiatry, 43, 809–811.

Beautrais, A., & Gibbs, S. (2004). Attempted suicide in Canterbury. Christchurch, New Zealand: Canterbury Suicide Project, Christchurch School of Medicine and Health Services.

Centers for Disease Control and Prevention (CDC). (2012). Deaths: Preliminary Data for 2010. Retrieved February 29, 2012, from http://www.cdc.gov/nchs/data/nvsr/nvsr60/nvsr60_04.pdf

Fleischman, A., Bertolote, J., Wasserman, D., DeLeo, D., Bolhari, J., Botega, N., et al. (2008). Effectiveness of brief intervention and contact for suicide attempters: A randomized controlled trial in five countries. Bulletin of the World Health Organization, 86, 703–709.

Fleiss, J.L. (1981.) Statistical Methods for Rates and Proportions, Second Edition. New York: John Wiley & Sons.

Gould, M.S., Kalafat, J., Harris–Munfakh, J.L., and Kleinman, M. (2007.) An evaluation of crisis hotline outcomes part 2: Suicidal callers. Suicide and Life–Threatening Behavior, 37(3), 338–352.

Gould, M.S., Munfakh, J.L., Kleinman, M., & Lake, A.M. (2012). National Suicide Prevention Lifeline: Enhancing mental health care for suicidal individuals and other people in crisis. Suicide and Life–Threatening Behavior, 42(1), 22–35.

Kalafat, J., Gould, M.S., Harris–Munfakh, J.L., and Kleinman, M. (2007.) An evaluation of crisis hotline outcomes part 1: Nonsuicidal crisis callers. Suicide and Life–Threatening Behavior, 37(3), 322–337.

Knesper, D.J., American Association of Suicidology, & Suicide Prevention Resource Center. (2010). Continuity of care for suicide prevention and research: Suicide attempts and suicide deaths subsequent to discharge from the emergency department or psychiatry inpatient unit. Newton, MA: Education Development Center.

Larkin, G. L., Smith, R., & Beautrais, A. (2008). Trends in US emergency department visits for suicide attempts, 1992–2001. Crisis: The Journal of Crisis Intervention and Suicide Prevention, 29, 73–80.

Motto, J. A., & Bostrom, A. G. (2001). A randomized controlled trial of postcrisis suicide prevention. Psychiatric Services, 52(6), 828–833.

Qin, P., & Nordentoft, M. (2005). Suicide risk in relation to psychiatric hospitalization: Evidence based on longitudinal registers. Archives of General Psychiatry, 62, 427–432.

Vaiva, G., Ducrocq, F., Meyer, P., Mathieu, D., Phillipe, A., Libersa, C., & Goudemand, M. (2006). Effect of telephone contact on further suicide attempts in patients discharged from an emergency department: randomised controlled study. BMJ, 332(7552).

List of Attachments

Attachment A	Hospital Data Abstraction Form
Attachment B	Crisis Center Data Abstraction Form (revised)

1 The date of admission (index admission) refers to each patient’s first admission within the time-frame of the current evaluation (the pre-collaboration or collaboration period). Some of the patients admitted for suicidal behavior during the study period may have prior admissions (prior to the evaluation) for suicidal behavior. Whether or not a patient had a prior admission (within the 365 days prior to the index admission) will be analyzed as a possible covariate of subsequent readmissions (within the 365 days following the index admission).

Page 6