Table 1.
Category | Point estimate | Range | Source |
---|---|---|---|
| |||
PREVALENCE | |||
| |||
Suicide risk status at time of initial ED presentation | |||
| |||
High risk | 2.8% | * | |
| |||
Low risk | 9.6% | * | |
| |||
No risk | 87.6% | * | |
| |||
Triage | |||
| |||
Medical Branch | 93% | 90% – 97% | * |
| |||
Percentage of the hospitalized patients among the general ED population (i.e., the study cohort) | 13.4% | 12.2%–14.6% | 38,39,** |
| |||
Of these, hospitalized for medical reasons, no sign of suicidality | 7.04%* | * | |
| |||
Of these, no risk | 100% | 100% – 100% | * |
| |||
Of these, hospitalized for medical reasons, apparent self-injury | 4.5% | 2% – 8% | ** |
| |||
Of these, high risk | 20% | 15% – 25% | ** |
Of these, low risk | 30% | 25% – 35% | ** |
Of these, no risk | 50%** | * | |
| |||
Of these, medical treatment in ED | 88.46% | * | |
| |||
Of these, high risk | 1.75% | 0.75% – 2.75% | 40,*** |
Of these, low risk | 8% | 6% – 10% | 27,40,*** |
Of these, no risk | 90.25% | 87.25% – 93.25% | 30,*** |
| |||
Psych Branch | 7% | 3% – 10% | 41 |
| |||
Of these, high risk | 7.5% | 5% – 10% | 41,*** |
Of these, low risk | 25% | 20% – 30% | 41,*** |
Of these, no risk | 67.5% | 60% – 75% | 30,*** |
| |||
SENSITIVITY & SPECIFICITY | |||
| |||
Medical Branch, suicide screening | |||
| |||
Sensitivity, high risk | 30% | 20%–40% | ** |
| |||
Sensitivity, low risk | 3% | 0% – 6% | ** |
| |||
Specificity, no risk | 99% | 95% – 100% | ** |
| |||
Medical Branch, suicide risk assessment among those with positive suicide screening | |||
| |||
Sensitivity, high risk | 95% | 90% – 100% | ** |
| |||
Sensitivity, low risk | 66% | 50% –80% | ** |
| |||
Specificity, no risk | 50% | 40% – 60% | ** |
| |||
Psych Branch, suicide risk assessment (100% assumed to have positive suicide screening) | |||
| |||
Sensitivity, high risk | 95% | 93%–97% | ** |
| |||
Sensitivity, low risk | 66% | 50%–80% | ** |
| |||
Specificity, no risk | 56% | 46% – 66% | 42 |
Sensitivity & specificity of identifying suicide risk in patients admitted to hospital from ED for medical reasons | |||
| |||
Sensitivity, high risk | 100% | 100% – 100% | ** |
| |||
Sensitivity, low risk | 100% | 100% – 100% | ** |
| |||
Specificity, no risk | 50% | 40% – 60% | ** |
| |||
EVENT PROBABILITIES – PSYCHIATRIC HOSPITALIZATION | |||
| |||
Medical Branch, positive suicide screen & positive suicide assessment | 35%* | 25% – 45% | 24,*** |
| |||
Psych Branch, positive suicide assessment | 80% | 70% – 90% | 24,*** |
| |||
Psych Branch, negative suicide assessment | 10% | 5% – 15% | ** |
| |||
BASE COSTS | |||
| |||
Medical ED visit (no risk; discharged alive) | $675 | $25–$2,850 | 24,*** |
| |||
Medical ED visit (high risk or low risk; discharged alive) | $890 | $25–$3,350 | 24,*** |
| |||
Psych ED visit (discharged alive) | $695 | $25–$2,950 | 24,*** |
| |||
Suicide risk assessment | $150 | $100–$200 | ** |
| |||
Medical hospitalization (no risk; discharged alive) | $8,765 | $1,450–$33,500 | 24,*** |
| |||
Medical hospitalization (no risk; died in the hospital) | $21,740 | $1,650 – $104,000 | 24,*** |
| |||
Medical hospitalization (high risk or low risk; discharged alive) | $11,080 | $1,550–$49,500 | 24,*** |
| |||
Medical hospitalization (high risk or low risk; died in the hospital) | $21,460 | $2,650–$123,000 | 24,*** |
| |||
Psychiatric hospitalization (discharged alive) | $5,875 | $1,050–$20,500 | 24,*** |
| |||
Psychiatric hospitalization (died in the hospital by suicide) | $18,790 | $1,350–$85,000 | 24,*** |
| |||
Inpatient suicide Tx | $2,000 | ** | |
| |||
DEATH AND REATTEMPT RATES | |||
| |||
Death by non-suicide manner (in 6 weeks) – same for all risk categories & treatments: Based on average non-suicide mortality for ages 35–44, US, general population, men & women combined, 2007–2010 | 0.02044% | 43 | |
| |||
Probability for a new suicide attempt (assuming no treatment) | |||
| |||
High risk –1st Markov cycle after the index event | 0.048 | 44,*** | |
| |||
High risk – 2nd – 4th Markov cycle, each | 0.00038 | 44,** | |
| |||
High risk – 5th Markov cycle | 0.00029 | 44,** | |
| |||
High risk – 6th – 9th Markov cycle, each | 0.00020 | 44,** | |
| |||
Low risk – no treatment (false negative) – distributed by 6-week cycles | 50% of high rate | ** | |
| |||
No risk | 0% | ** | |
| |||
Ratio of suicides to suicide attempts | 1:13 | 45,46 | |
| |||
Years of Potential Life Lost per suicide | 24 | 46,*** | |
| |||
INTERVENTIONS - UPTAKE, OUTCOMES and COSTS | |||
| |||
Usual Care (UC; also provided to people who receive inpatient suicide treatment)*** | |||
| |||
Uptake (i.e., this % receive any outpatient suicide treatment) | 35% | 10% – 50% | ** |
| |||
Reduction in (re-)attempt rate, vs. no treatment | 15% | 10% – 20% | ** |
| |||
Cost [based on CPT 90791 (psychiatric diagnostic evaluation) + 2 times CPT 90834 (45 min psychotherapy)] | $340 | 23,** | |
| |||
Postcards (PC)* | |||
| |||
Uptake | 100% | NA | 11 |
| |||
Reduction in (re-)attempt rate, vs. UC | 45% | 35% – 55% | 11,** |
| |||
Additional cost of intervention [based on $10 per person for the postcards per se; plus $135 (1.5 CPT 90834 visits) in additional outpatient treatment as function of receiving the postcards] | $145 | $135 – $500 | 11,** |
| |||
Telephone Outreach (TO) | |||
| |||
Uptake (i.e., this % of those offered TO participate; the rest do UC only) | 70% | 60% – 80% | 12,** |
| |||
Reduction in (re-)attempt rate among those with TO uptake, vs. UC | 34% | 25% – 45% | 12,** |
| |||
Additional cost of intervention, for those with uptake [based on $30 for the phone calls per se, plus $270 (3 CPT 90834 visits) in additional outpatient Tx as function of receiving the calls] | $300 | $300 – $900 | 12,** |
| |||
Cognitive Behavioral Therapy (CBT) | |||
| |||
Uptake (i.e., this % of those offered CBT participate; the rest do UC only) | 65% | 55% – 75% | 13,** |
| |||
Reduction in (re-)attempt rate among those with uptake, vs. UC | 50% | 40% – 60% | 13,** |
| |||
Additional cost of intervention, for those with uptake [based on 9 times CPT 90834] | $810 | $810 – $2000 | 13,*** |
NOTES:
Calculated by the authors based on available data regarding prevalence of suicide risk within subgroups of the ED population (i.e., the sources cited under “Triage” in Table 1), so that each individual presenting to the ED for the initial (index) visit is counted exactly once
Author opinion
Author opinion, considering information available from the listed source(s)