Identification of Seniors At Risk

Adults 65+ years old: (Singler et al., 2014; n=520 German adults in ED)

• At 28 days
  • ISAR >=2 : 24.8% (low specificity), 88.8% (high sensitivity)
  • ISAR >=3 : 49.3% (moderate specificity), 71.2% (moderate sensitivity)
• At 180 days
  • ISAR >=2 : 27.3%(low specificity), 90.1% (high sensitivity)
  • ISAR >=3 : 52.7% (moderate specificity), 73.8% (moderate sensitivity)

• Despite high sensitivity, low specificity was seen by this author as a reason to choose the ISAR >= 3 to decrease the number of false positive screenings for the combined endpoint.

 (Dendukuri et al., 2004; Study 1 n=1,122, Study 2 n=1,889)

• Cut-off point of 2 had greater sensitivity in predicting severe functional impairment, detecting depression, detecting increased depressive symptoms, and frequent use of community health centers.

(Suffoletto et al., 2016; n=202 United States adults)

• The optimum ISAR cut-off score for screening was >=2 with high sensitivity of 91% and low specificity of 19%

Patients with Previous Mobility Deficits: (Singler et al., 2014)

• ISAR score 0-1 (n=95): 8.4% with previous mobility deficits
•  ISAR score >=2 (n=425): 50.6% with previous mobility deficits

(Salvi et al., 2009; n=200 Italian adults)

• Katz ADL score for ISAR 0-1: mean 5.5 (1.1)
• Katz ADL score for ISAR ³2: mean 3 (2.3)

Fall History: (Edmans et al., 2013; n=667 UK adults)

• ISAR score 0-1 (median age 78; n= 205): 23.4% presented with falls
•  ISAR score >=2 (median age 81; n=462): 24.5% presented with falls

Using >= 3 medications: (Dendukuri et al., 2004)

• ISAR score 0-1: Study 1-(n=608): 17.8%; Study 2- (n=132): 25%
• ISAR score 2-6: Study 1-(n=514): 69.1%; Study 2- (n=288): 79.1%

Admitted to Hospital from ED visit: (Salvi et al., 2009)

• ISAR score 0-1 (median age 78.3 (7.5); n= 59)- 44% admitted to hospital
•  ISAR score >=2 (median age 81.2 (7.2); n=141)-63.8% admitted to hospital

(Asomaning &Loftus, 2014; n=258; United States adults)

• ISAR score negative [ie <=2] (n=100)- 23% admitted to hospital
• ISAR score  positive[ie >=2] (n=158)- 50% admitted to hospital

Gender: (Singler et al., 2014)

• ISAR score >=2: 62.6% were women

(Edmans et al., 2013)

• ISAR score >=2: 60.4% were women

Age: (Asomaning &Loftus, 2014)

• ISAR score negative [ie <=2](n=100): mean age 76.6 (7.5)
• ISAR score  positive[ie >=2] (n=158): mean age 80.9 (8.5)

(Salvi et. al, 2009)

• ISAR score >= 2: mean age 81.2 (7.2)
• ISAR score 0-1: mean age 78.3 (7.5)

(Suffoletto et al., 2016)

• Mean age of those who reached the combined endpoint of (ED visit, post discharge hospitalization or death with 30 days of initial ED visit) was 77 (7.8)

Adults 65+ years old in an ED: (McCusker et al., 1998; n=404)

• Excellent test-retest reliability of the total risk factor score (ICC=.78)

Predictive validity:

Adults 65+ years old: (Rosted et al., 2004; n=278; mean age=78 years)

• Adequate correlation found between patient’s ISAR score and the number of days spent in the hospital (Pearson’s r=0.36, p=0.00)and their number of health problems (Pearson’s r=0.55, p=0.00)
• Adequate correlation found between the ISAR score and readmissions within one (g=0.32; p=0.02) and three months (g=0.51; p=0.00)

 (Buurmann et al., 2011, n=381; performed in a Netherlands EDt)

• Poor correlation with death- AUC: 0.58
• Poor correlation with hospital admission- AUC: 0.59
• Poor correlation with recurrent visit to ED- AUC: 0.59
• Poor correlation with poor outcomes (combined)-AUC: 0.60

(Salvi et al., 2012)

• Poor correlation with return visit to ED-AUC: 0.61
• Poor correlation with hospital admission-AUC: 0.61
• Adequate correlation with death-AUC: 0.75

(Dendukuri et al., 2004)

• Adequate correlation with severe functional impairment-AUC: 0.86 (both studies)
• Poor correlation with frequent ED visits-Study 1 AUC: 0.61, Study 2 AUC: 0.68
• Poor correlation with frequent hospitalization- Study 1 AUC: 0.66, Study 2 AUC: 0.68 
• For increased depressive symptoms: moderate sensitivity 73% and specificity 57% (Study 2)
• For frequent emergency room usage (>=2 visits): moderate sensitivity 71% and specificity 60% (Study 2)
• For frequent hospitalization (>=11days): moderate sensitivity 70% and specificity 58% (Study 2)
• For frequent use of community health center (10 visits): moderate sensitivity 73% and specificity 59% (Study 2) 
• Ability of the questions to predict severe functional impairment: high sensitivity-94%, moderate specificity-58%
• Ability of the questions to predict depression: high sensitivity-83%, moderate specificity-63%

(Singler et al., 2014)

• Poor correlation for combined endpoint on day 28- AUC: 0.621
• Poor correlation for combined endpoint on day 180- AUC: 0.661
• RR (relative risk):
  • ISAR >=2 : 1.77 (28 days), 2.20 (180 days) increased risk
  • ISAR >=3 : 1.61 (28 days), 3.14 (180 days) increased risk

(Edmans et al., 2013; cutoff score>=2)

• Moderate Sensitivity- ISAR >=2 : 71% (at 90 day follow-up)
• Low Specificity- ISAR >=2 : 43% (at 90 day follow-up)

 (Buurman et al., 2011; ISAR cutoff score >=2)

• Moderate Sensitivity- 65% (at 120 days follow-up)
• ModerateSpecificity- 54% (at 120 days follow-up)

 (McCusker et al., 1999; n=1,854; Cut-off score ISAR >=2)

• Moderate Sensitivity- 72% (at 6 month follow-up)
• Moderate Specificity- 58% (at 6 month follow-up)
• Adequate correlation with adverse outcome- AUC=0.71

(McCusker et al., 1998)

• The questions with the highest level of both sensitivity and specificity were those on visual and hearing impairment as well as those on medications (sensitivity >=60% and specificity >=90%)

(Salvi et al., 2009)

• All items of the ISAR were able to predict the composite outcome (OR- >2.2 and r below 0.014) except polypharmacy (OR-1.72, r =0.20)

(Suffoletto et al., 2016)

• The ISAR risk factor with the best prognostic accuracy of poor outcome at 30 days was the concurrent use of >= 3 medications with a high sensitivity of 93% and a low specificity of 13%

(Salvi et al., 2009)

• High Adjusted OR (adjusted odds ratio) for composite outcome=4.85, r<0.0001
• High Adjusted OR (adjusted odds ratio) for frequent ED return=4.69, r<0.0019
• Moderate Adjusted OR (adjusted odds ratio) for functional decline=2.98, r=0.016
• Moderate Adjusted OR (adjusted odds ratio) for admission to hospital=2.07, r=0.043
• Low Adjusted OR (adjusted odds ratio) for early ED revisit=1.67, r=0.20
• High Hazard Ratio- 6.9

Adults 70+ years old in an acute medical unit: (Edmans, 2013)

• Poor correlation with death- AUC :0.62
  • PPV-6%, NPV-97%
• Poor correlation with move to care home-AUC: 0.65
  • PPV-1%, NPV-99%
• Poor correlation with readmission- AUC:0.60
  • PPV-29%, NPV-79%
• Poor correlation with increase in dependency- AUC :0.62
  • PPV-24%, NPV-88%
• Poor correlation with reduced mental well being- AUC :0.50
  • PPV-45%, NPV-54%
• Poor correlation with reduced quality of life- AUC: 0.56
  • PPV-53%, NPV-60%
• Poor correlation with any adverse outcome- AUC: 0.60
  • PPV-79%, NPV-32%
• Adequate correlation with high total health and social care costs-AUC :0.70
  • PPV-13%, NPV-96%

Adults 65+ years old in an ED: (Salvi et al., 2012; n=2,057; mean age 81.7)

• Adequate correlation between ISAR and Triage Risk Screening Tool (TRST): Pearson r=0.539

• Items were developed from a literature review of general factors and specific questions that predict functional decline
• Potential screening questions were sent to multidisciplinary hospital and community-based health professionals who were asked to prioritize questions addressing each topic of interest
• These test items were then pretested on a sample of ED patients to determine ease of comprehension
• Twenty-seven questions were initially used.  Those questions were then studied to determine which screening questions were the best subset for predicting a disability at 6 months later
• This 27 item questionnaire was then refined to 6 items