Alex Mitchell www.psycho-oncology.info
Paul SymondsLorraine GraingerElena Baker-Glenn
Department of Cancer & Molecular Medicine, Leicester Royal Infirmary
IPOS 2010
T276 - A Large Scale Validation of the Emotion Thermometers as a Screening Tool for MDD & Distress in an Ethnically Diverse Cancer Population
T276 - A Large Scale Validation of the Emotion Thermometers as a Screening Tool for MDD & Distress in an Ethnically Diverse Cancer Population
Audit / Research Protocol
Phase I – DT across LNR network (incl training)
Phase II – Enhancements to DT
Phase III - Screening plus Intervention
Qualitative Limitations of DT
DT not always interpreted by patients
Distress not always understood by patients
There is no anchor
Patients who are angry or depressed may not say “distressed”
But some patients may not interpret “depression”
Quantitative Limitations of DT
Of 401 chemotherapy attendees
59% have an emotional complication (3v4)
37% (62% of 59%) it included distress
23% it excluded distress
Validity of DT vs depression (DSMIV)
SE 80% SP 60% PPV 32% NPV 93%
ET - Table of Cut-Points
11.2
6.6
12.2
15.4
54.3
Help Thermometer
8,9,1013.69.520.811.9Severe
6,76.69.814.512.0Moderate
4,510.512.216.516.9Mild
2,313.618.322.520.1Minimal
0,155.750.125.639.0Insignificant
Cut-point
Anger Thermometer
Depression Thermometer
Anxiety thermometer
Distress Thermometer
p130
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
DistressThermometer
AnxietyThermometer
DepressionThermometer
AngerThermometer
TenNineEightSevenSixFiveFourThreeTwoOneZero
8%
DT37%
DepT23%
AngT18%
AnxT47%
4%
7%
1%
1%
9%
3%
0%
2%
4%
15%
3%
2%
Nil41%
Non-Nil59%
DT
AnxT AngT
DepT
Top 10 patient concerns
49Breathing1052Weight963Finances871Fatigue/Energy776Sleep686Pain5
111Appearance4
122Independence/Role3
157Family concerns2
312Anxiety1
Concerns linked with distress
SleepP = 0.0183t = 2.371534
Coping concernsP = 0.013t = 2.497361
Cancer worriesP = 0.0073t = 2.69652
Self-EsteemP = 0.0073t = 2.699969
Anger/IrritabilityP < 0.0001t = 4.10916
Family concerns/issuesP < 0.0001t = 4.448083
depression/hopelessP < 0.0001t = 4.859186
AnxietyP < 0.0001t = 7.705129
Graphical – Screening principles
Non-Depressed
Depressed
# ofIndividuals
Cut-Off
# ofIndividuals
Severity of Depression
HighLow
High Sensitivity >>>>
<<<< low Specificity
Validity Against HADS-T (distress)
We analysed data collected from Leicester Cancer Centre from 2007-2009
The researcher applied the HADS and used a HADS-T >14 to signify distress.
We collated full data on 660 patient assessments
HADS-T
In the parent sample of 660:
SE SP AUC CUTDT – 71.9% 78.4% 0.814 cut point =4
AnxT – 75.7% 73.4% 0.821 cut point =5
DepT – 77.6% 82.2% 0.855 cut point =3
AngT – 77.5% 77.6% 0.823 cut point =2
HelpT - 69.1% 80.8% 0.809 cut point =3
Against MDD
The researcher applied criteria for MDD. We collated full data on 660 patient assessments
MDD was defined by DSMIV criteria (5 of 9 symptoms)
12.9% had MDD and 14.8% were from ethnic minorities (largely British South Asian of India descent).
Validity vs MDD - Results
SENSITIVITY SPECIFICITY AUC
DT – 82.4% 68.6% 0.811
AnxT – 85.9% 56.2% 0.774
DepT – 80.0% 78.2% 0.853
AngT – 83.5% 66.1% 0.782
HelpT - 68.2% 79.1% 0.799
Whole Sample vs MDD
0.00%
10.00%
20.00%
30.00%
40.00%
50.00%
60.00%
70.00%
80.00%
90.00%
100.00%
DT – AnxT – DepT – AngT – HelpT -
SensitivitySpecificity
Validity vs MDD – 15% minorities
SENSITIVITY SPECIFICITY AUC
DT BSA - 100% 56.9% 0.827
AnxT BSA - 72% 72.1% 0.730
DepT BSA - 80.0% 69.4% 0.770
AngT BSA - 84.0% 65.3% 0.782
HelpT BSA - 92% 56.9% 0.791
BSA vs MDD
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
DT – AnxT – DepT – AngT – HelpT -
SensitivitySpecificity