14 Journal of The Association of Physicians of India ■ Vol.

64 ■ February 2016

Original Article

Towards Developing a Scoring System for

Febrile Thrombocytopenia
Prasita Kshirsagar1, Shaylika Chauhan2, Dinesh Samel3

Abstract Editorial Viewpoint

Background: The authors wished to develop a scoring system for • F e v e r w i t h
evaluating patients presenting with febrile thrombocytopenia for risk thrombocytopenia is a
stratification, predicting patient outcome and optimization of care common occurrence.
especially in resource poor countries. • Calculation of total risk
Method: Objective: 1. To decide a protocol in the management of patients score based on platelet
with fever and thrombocytopenia. 2. To develop screening or therapeutic count, vital signs and
guidelines (early warning score-EWS) in febrile thrombocytopenic systemic complications is
attempted here.
patients and decide about therapeutic interventions
• Total risk score appears
Design: Retrospective study and development of a bedside scoring system
to be better indicator for
based on Platelet Count, Temperature, Respiratory Rate, Blood Pressure.
platelet transfusion than
Pulse, CNS, Respiratory, Hematological, Hepatic and Renal complications
platelet counts.
in a central civic hospital and teaching institute in India
• Further validation of this
Par ticipants: All patients >18 years presenting with fever and scoring system is needed.
thrombocytopenia with platelet count of <150 × 109/L.
Results: Number of patients requiring platelet transfusions decreases of patients in the ICU setting
when total risk score is used for risk stratification and for transfusing were thrombocytopenic in a study
platelets as against the platelet count at admission. Patients who died carried out by Teo et al. 2 Severe
in our study had a platelet count at presentation between 20,000- thrombocytopenia can lead to
1,00,000 though their total risk score was 17 and 18 respectively; hence life-threatening complications like
platelet count alone should not be relied upon for platelet transfusion. intracerebral or intra-abdominal
Irrespective of the number of platelets transfused the prognosis is poor bleeding.
as the total risk score increases. However an issue which has
Conclusion: The platelet count is not the only indicator of transfusion. n o t b e e n r e s o l ve d i s w h e t h e r
When we use total risk score instead of platelet count for classifying platelet count alone should be
patients who need transfusions, number of patients who fall in severe risk relied upon to initiate a platelet
category needing immediate transfusion reduces and haphazard use of transfusion in acute medical
platelets can be avoided. Patient outcome (death/survival), occurrence of settings and the number of platelets
to be transfused. Also definitions
complications and hematological manifestations (petechiae/purpura etc)
o f m i l d , m o d e r a t e a n d s e ve r e
are not dependent on platelet count at presentation. There is a significant
thrombocytopenia do not exist. As
association between risk category and patient outcome.
per Council of Europe Guide to
the Preparation, Use and Quality
Assurance of Blood Components
Introduction Examination Survey (NHANES the decision to transfuse platelets
III) as the lower limit of normal. 1 must not be based exclusively on

H istorically thrombocytopenia
is defined as a platelet
count below the normal range
About 0.9% of patients in the
acute care setting and 25 to 41%
the platelet count 3 and must take
patient’s clinical condition into

for the population of - 2 standard

deviations and the traditional Associate Professor, 2Clinical Research Assistant, Department of Medicine, 3Associate Professor, Department
1

of P.S.M., Rajiv Gandhi Medical College, Kalva, Maharashtra

value of 150 x 10 9 is supported by
Received: 21.08.2014; Accepted: 17.02.2015
the National Health and Nutrition

Table 1: Clinical scoring system for thrombocytopenia
Platelet Count
Blood pressure
Pulse
Respiratory Rate
Temperature
CNS complications*
Respiratory complications*
Haematological complications*
Hepatic complications*
Renal complications*
*
Zero if absent
Table 2: Platelet count and other parameters at admission and number of platelet
packs given (N=93)
Platelet count
Mean 46225.81
Median 39000.00
Std. Deviation 29108.167
account. However platelets are
transfused based on platelet counts
alone due to lack of guidelines.
Overzealous use of platelet
transfusions overwhelms blood
banks jeopardising their supply in
essential cases. As demonstrated by
various studies thrombocytopenia
is an early warning sign of sepsis
and portends poor prognosis. 4
Keeping these factors in mind,
we have described a simple
scoring system (total risk score) for
management of thrombocytopenia
in acute clinical conditions.
Aim
1 . T o de c i de a p rotoc ol in th e
management of patients with
fever and thrombocytopenia
2. To develop screening or
therapeutic guidelines (early
warning score-EWS) in febrile
thrombocytopenic patients
and decide about therapeutic
interventions in the form of
platelet transfusions.
Material and Methods
A retrospective study spanning
over a period of four months from
June 2010-September 2010 was
carried out in Rajiv Gandhi Medical
College, Thane, Maharashtra.
A total of 93 subjects admitted
to the medical ward or ICCU with
fever and thrombocytopenia were
Journal of The Association of Physicians of India ■ Vol. 64 ■ February 2016 15
Table 3: Test result variable(s): total
risk score
1 2 3
>1,00,000 20,000-1,00,000 <20,000 Positive if ≥ Sensitivity 1-Specificity
>90 ≤90 5.00 1.000 1.000
<100 ≥100 6.50 1.000 .912
<20 ≥20 7.50 1.000 .703
 <100      ≥100 8.50 1.000 .527
Present 9.50 1.000 .418
Present 10.50 1.000 .352
Present 11.50 1.000 .264
Present 12.50 1.000 .154
Present 13.50 1.000 .088
14.50 1.000 .044
16.00 1.000 .011
17.50 .500 .011
18.50 .000 .011
Platelets given Pulse rate Temp RR Systolic BP 20.00 .000 .000
2.09 92.58 99.146 20.40 109.29 The smallest cut-off value is the minimum
.00 90.00 98.500 21.00 110.00 observed test value minus 1, and the largest
2.858 13.167 1.4310 2.751 14.887 cut-off value is the maximum observed test
value plus 1. All the other cut-off values are
included in the study. All the study the averages of two consecutive ordered
subjects were identified to have observed test values.
thrombocytopenia defined as a
Statistical Analysis
platelet count of <150 × 10 9/L.
Statistical analysis was
Patients younger than 18
performed using the statistical
years, pregnant patients,
software package SPSS (version 20). 
patients on drugs causing
The non-parametric Chi-square
thrombocytopenia (chemotherapy/
test and t-test were also applied
immunosuppressants/
in comparative analysis results
antiplatelets), previously
between different groups and to
diagnosed patients with chronic
find significance (p) value. Mean
thrombocytopenia due to any cause
values, standard deviation,
were excluded.
prevalence was assessed wherever
Evaluation: Records of vital signs relevant.
of the patients at time of admission
were noted. Clinical examination Results
findings of patients with respect
to neurological status, respiratory, The mean age of patients
renal, haematological(petechiae, presenting with thrombocytopenia
purpurae, ecchymosis or bleeding was 32.2 ± 13.2. 83.9% patients were
male and 16.1 percent patients were
from any orifices) and other
female.
complications were noted. Routine
investigations such as complete Mean number of platelets at
blood count, plasma glucose, urea, a d m i s s i o n wa s 4 6 , 2 2 5 a n d t h e
serum creatinine, bicarbonate, mean number of bags of platelets
serum electrolytes, bilirubin, transfused was two (Table 2). There
alanine aminotransferase were were two deaths in the cohort.
noted. In some of the patients cause Patients were assigned a score of
of fever was found and tabulated 1, 2 or 3 depending on parameters
accordingly. as mentioned in Table 1. Range
The study protocol was approved of score as per scoring system is
by the Institutional Human Ethical 5 to 26 as per Table 1. As per our
committee (IHEc) of Rajiv Gandhi study minimum score was 6 and
Medical College. maximum score was 19.
Early warning score was Best Cut-off was 16 with
prepared as shown in Table 1. Sensitivity 100% and Specificity
16 Journal of The Association of Physicians of India ■ Vol. 64 ■ February 2016

Table 4: Frequency and percentage of Table 5: Frequency and percentage of per total risk score have a worse
platelet count total risk score group outcome (Table 7). Patients who
Platelet count Frequency (%) Total risk score Frequency (%) survived had a mean total risk
>100,000 4 (4.3) Low (6-7) 27 (29.0) score of 9.51 and those who died
20,000-100,000 69 (74.2) Moderate (8-15) 63 (67.7) had a score of 17.50.
<20,000 20 (21.5) High (16-26) 3 (3.2) The outcome of the patients is
Total 93 Total 93 dependent on the risk score and
Table 6: Table comparing outcome (survival vs death) with platelet count categories not only on the platelet count at
a d m i s s i o n . I r r e s p e c t i ve o f t h e
Platelet count
number of platelets transfused the
Outcome >100,000 20,000-100,000 <20,000 Total
prognosis is poor as the risk score
Survived 4 (100) 67 (97.1) 20 (100) 91 (97.8)
increases (Table 8).
Died 0 (0) 2 (2.9) 0 (0) 2 (2.2)
Total 4 (100) 69 (100) 20 (100) 93 (100) Tables 9-13 show that organ
system involvement which
No. of patients (%)
increases total risk score
Table 7: Table comparing outcome with total risk score can occur with any degree of
Outcome thrombocytopenia (statistically no
Risk group Survived Died Total significant difference).
Low (6 - 7) 27 (100) 0 (0) 27 (100 Mean platelet count at admission
Moderate (8-15) 63 (100) 0 (0) 63 (100) of patients with hematological
High (16-26) 1 (33.3) 2 (66.7) 3 (100) involvement was 35,888 and
Total 91 (97.8) 2 (2.2) 93 (100) without hematological involvement
No. of patients (%). 2 patients who died belonged to the high risk score category was 47,333. Table 14 shows that the
Table 8: Comparison of platelet at admission, no of platelet received and total risk
cut of <20,000 for indicating chance
score with the outcome of bleeding cannot be relied upon.

Outcome N Mean SD SE t-value df P

Platelet at Survived 91
admission Died 2 23500.0
No. of platelets Survived 91
received Died 2 5.50
Total risk score Survived 91
Died 2 17.50
9 8 . 9 % f o r p i c k i n g u p a d ve r s e
outcome of patients presenting
with febrile thrombocytopenia.
Hence 16 is taken as the cutoff for
putting patients in the high risk
category (Table 3).
For low risk category cut off is 7
as minimum score after considering
parameters of temp, RR, BP, pulse
and platelet count is 5. Presence of
any complication pushes score to
8 and hence a score of more than 7
is placed in moderate risk category
(Table 3).
4.3, 74.2 and 21.5% patients had
a platelet count >1,00,000, 20,000-
1,00,000 and <20,000 respectively
(Table 4). 29%, 67.7% and 3.2% of
patients were in low, moderate
and high risk category respectively
(Table 5).
As per the study done, in high
risk group with score >16 (group 3)
Pa t i e n t s r e c e i v i n g p l a t e l e t
46725.3 29227.5 3063.9 1.118 91 .267 transfusions were 16, 3, 6, 2
3535.5 2500.0 and 18 with renal, respiratory,
2.01 2.842 .298 5.994 91 .033 hematological, CNS and hepatic
.707 .500 system involvement (Table 15).
9.51 2.734 .287 4.112 91 .000
.707 .500
Discussion
all the patients received platelets, Thrombocytopenia is a
in moderate risk group 33 out of haematological finding which is
63 patients received transfusions commonly seen in clinical settings.
while in low risk score category It may be the manifestation of
2 patients out of 27 received common infectious diseases like
transfusions. dengue, malaria, leptospirosis or
1 patient with platelet count it may be due to life-threatening
of more than 1,00,000, 16 patients multisystem illnesses like DIC or
in group 20,000-1,00,000 and all thrombotic microangiopathies.
20 patients with platelet count Va r i o u s v i r a l i n f e c t i o n s ( e . g . ,
less than 20,000 were transfused rubella, mumps, varicella,
platelets. parvovirus, hepatitis C, and
Epstein-Barr virus), HIV may
Patients who died in our study
present with thrombocytopenia.
had a platelet count at presentation
Other illnesses associated with
between 20,000 - 1,00,000 though
febrile thrombocytopenia include
their total risk score was 17 and 18
meningococcemia, rat-bite fever,
respectively. There is no significant
rickettsial infections, hantavirus,
difference across groups of platelet
and other viral hemorrhagic fevers
count in patient outcome (Table 6).
(e.g., Ebola, Lassa fever ). 5
Since P<0.0001 there is a
Patients presenting with fever
significant association between
and thrombocytopenia may
risk category and outcome. Patients
require emergent interventions and
who fall in high risk category as
intensive care support. There is a
 Journal of The Association of Physicians of India ■ Vol. 64 ■ February 2016 17

Table 9: Hepatic complications and platelet count thrombocytopenia. Adverse outcomes generally
Hepatic complications correlate with the severity of thrombocytopenia as
Platelet count No Yes Total demonstrated by PROTECT trial which included a
>100,000 3 (4.5) 1 (3.8) 4 (4.3) cohort of medical and surgical ICCU patients; however
20,000-100,000 53 (79.1) 16 (61.5) 69 (74.2) such a study has not been carried out in patients
<20,000 11 (16.4) 9 (34.6) 20 (21.5) presenting with febrile thrombocytopenia. 9 The
Total 67 (100) 26 (1000 93 (100) only scoring system available for thrombocytopenia
Table 10: Renal complications and platelet count is for heparin-induced thrombocytopenia and
there is a paucity of research in evaluating febrile
Renal complications
thrombocytopenia. 10 APACHE-II scoring system can
Platelet count No Yes Total
be applied to these patients but it’s too cumbersome
>100,000 4 (5.8) 0 (0) 4 (4.3)
and difficult to apply for bedside evaluation of the
20,000-100,000 49 (71) 20 (83.3) 69 (74.2)
patient. 11 We have tried to develop a scoring system
<20,000 16 (23.2) 4 (16.7) 20 (21.5)
which is free of interobserver variability, depends on
Total 69 (100) 24 (100) 93 (100)
few bedside variables and minimal lab parameters
Table 11: Respiratory complications and platelet count and is easy to remember. Patients presenting with
Respiratory complications febrile thrombocytopenia can be classified into low,
Platelet count No Yes Total intermediate and high risk depending on their score
>100,000 4 (4.4) 0 (0) 4 (4.3) and accordingly decisions about their management,
20,000-100,000 67 (74.4) 2 (66.7) 69 (74.2) daily prognosis assessment and referral to higher
<20,000 19 (21.1) 1 (33.3) 20 (21.5) centres can be made.
Total 90 (100) 3 (100) 93 (100)
Platelet transfusion is indicated in all patients with
Table 12: Hematological complications and platelet count platelet count less than 20,000 and otherwise stable,
Bleeding complications nonbleeding, and body temperature >100.4°F (38°C)
Platelet count No Yes Total or undergoing invasive procedure and platelet count
>100,000 3 (3.6) 1 (11.1) 4 (4.3) <10,000 who are otherwise stable and nonbleeding. 12
20,000-100,000 64 (76.2) 5 (55.6) 69 (74.2) Our study is done to find out the indication for
<20,000 17 (20.2) 3 (33.3) 20 (21.5) platelet transfusions in patients with platelet count
Total 84 (100) 9 (100) 93 (100) more than 20,000 who can have a poor prognosis and
Table 13: CNS complications and platelet count may require timely intervention. So we have graded the
various clinical parameters to get a scoring system and
CNS consciousness
rationalise the indications for platelet transfusion. The
Platelet count No Yes Total
>100,000 4 (4.4) 0 (0) 4 (4.3)
total risk score can be monitored frequently at bedside
20,000-100,000 67 (73.6) 2 (100) 69 (74.2)
of patient to provide ongoing care.
<20,000 20 (22) 0 (0) 20 (21.5) The scoring system will help in deciding about the
Total 91 (100) 2 (100) 93 (100) urgency of transfusing platelets and number of platelets
Table 14: Platelets at admission and bleeding complications to be transfused.

Bleeding N Mean SD SE
If we depend on platelet count alone for transfusing
complications patients as per standard recommendations, only four
Platelet at No 84 47333.3 29145.0 3180.0 patients fell in low risk category who did not require
admission Yes 9 35889.0 28246.4 9415.5 platelet transfusions but were 27 patients in low risk
Table 15: Organ system complications and platelet transfusion category if we take a risk score under consideration for
classifying patient risk. Hence the number of patients
Organ system No. of No. of Mean no. of Eta not requiring platelet transfusions is increased when
complications patients platelet platelet bags Squared
receipients transfused total risk score is used for risk stratification and for
Renal 24 16 3.37 0.125 transfusing platelets.
Respiratory 3 3 7.67 0.129 If we depend on score group for classifying patients,
Hematological 9 6 3.56 0.029 only two patients fell in the severe category in whom
CNS 2 2 5.50 0.032 transfusion is mandatory. If we depend on platelet
Hepatic 26 18 3.73 0.130 count alone for transfusion, then 20 patients fell in
need to develop a scoring system for thrombocytopenia this category for mandatory transfusion. There is a
which can predict case mortality, length of hospital definite decrease in the number of patients who require
stay and need of emergency interventions. There is a transfusion if we rely on platelet count alone.
scoring system in place for many illnesses like stroke, 6 Death occurred in patients irrespective of platelet
malaria, 7 and cardiac failure; 8 however there is no count category which again proves the point that
system to evaluate patients presenting with febrile
18 Journal of The Association of Physicians of India ■ Vol. 64 ■ February 2016
thrombocytopenia is not the only
indication for poor prognosis of
patient.
In our study platelet transfusion
was given to patients in a haphazard
manner irrespective of the degree
of severity of thrombocytopenia.
The platelet count and the
total risk score comparison for
transfusing platelets reveals that
the platelet count is not the only
indicator of transfusion. When
we use total risk score instead
of platelet count for classifying
patients who need transfusions,
number of patients who fall in
severe risk category needing
immediate transfusion reduces
and haphazard use of platelets can
be avoided.
In our study 0.09% of the
study population had bleeding
manifestations and these symptoms
were not dependent on platelet
count, hence all cases need not
present with haematological
manifestations. Since a patient
presenting with a higher score
had significantly worse outcome
including death these patients
should be adequately monitored
and provided ICU care if necessary.
Platelet count at presentation was
not related to poor prognosis. As
per our study patients who died
presented with a platelet count of
more than >20,000 and one of the
two patients who died didn’t have
any haematological manifestations.
Hence platelet count or bleeding
alone should not be relied on for
gauging prognosis of patients with
febrile thrombocytopenia. Similarly
an otherwise stable patient with a
low score in spite of presenting
wi t h se ve r e t h r ombocytop enia
can be closely monitored without
platelet transfusion.
The dynamic and erratic course
of thrombocytopenia requires
close monitoring of the patient
and careful decision making. The
scoring system developed by the
authors may serve as reproducible,
i n e x p e n s i ve b e d s i d e t o o l s f o r
evaluating thrombocytopenia and
in guiding platelet transfusion
therapy. It will help clinicians to
be more judicious in transfusing
platelets and it will guide early
i n t e r ve n t i o n s i n p a t i e n t s w i t h
a higher score so that these
patients can make a complete
recovery. Since this score has been
developed on simple physiological
parameters besides platelet count
at presentation it will help in
predicting case recovery and
survival. This scoring system can
fulfil the lacunae in evaluating
febrile patients presenting with
thrombocytopenia.
The scores can be a guide to the
management as below.
Low Risk Score (upto 7)
No platelet transfusion required.
No admission required.
Monitor the platelet count.
# Admission will depend upon
feasibility of monitoring platelet
count.
Moderate Risk Score (8-15)
Admit the patient in ward.
Consider transfusion of platelets
after monitoring
Revise the scoring system every
2-4 hours.
Severe Risk Score (16 and above)
Admit in ICCU
Transfuse platelets immediately
Limitation of Study
Study was retrospective so no
active intervention has been carried
out.
References
1. Cheng CK, Chan J, Cembrowski GS, and van
Assendelft  OW. Complete blood count
reference interval diagrams derived from
NHANES III: stratification by age, sex, and
race. Lab Hematol 2004; 10:42–53.
2. Te o C P, K u e h Y K . I n c i d e n c e o f
thrombocytopenia in an acute carehospital.
Ann Acad Med Singapore 1989; 18:379–381.
3. Liumbruno G, Bennardello F, Lattanzio A.
Recommendations for the transfusion of
plasma and platelets. Recommendations
for the transfusion of plasma and platelets.
Blood Transfus  2009; 7:132–150.
4. Moreau D. Platelet count decline: An Early
Prognostic Marker in Critically Ill Patients
with Prolonged ICU stays. Chest 2007;
131:1735-41.
5. Sekhon SS, Roy V. Southern Medical Journal
2006; 99:5.
6. Walter BG, Robert DB, Gerald FM, Joseph
J. Neurology in clinical practice, principles of
diagnosis and management. Philadelphia
Elsevier, 2004.
7. Mohapatra MK, Das SP. The malaria severity
score: a method for severity assessment
and risk prediction of hospital mortality
for falciparum malaria in adults. J Assoc
Physicians India 2009; 57:119-26.
8. Howard J. Eisen, the MELD Scoring System
and the Prediction of Outcomes in Heart
Failure Patients. J Am Coll Cardiol 2013;
61:2262-2263.
9. Cook D, Meade M, Guyatt G, et al.
PROphylaxis for ThromboEmbolism Critical
Care Trial protocol and analysis plan. J Crit
Care 2011; 26:223.e1-223.e9.
10.  Warkentin TE, Heddle NM. Laboratory
diagnosis of immune heparin-induced
thrombocytopenia. Curr Hematol Rep 2003;
2:148–57.
11. Knaus WA, Draper EA, Wagner DP,
Zimmerman JE. APACHE II: a severity of
disease classification system. Critical Care
Medicine 1985; 13:818–29.
12.  Liumbruno G, Bennardello F, Lattanzio
A , P i c c o l i P, R o s s e t t i G . I t a l i a n
Society of Transfusion Medicine and
Immunohaematology (SIMTI) Work Group.
Recommendations for the transfusion of
plasma and platelets. Blood Transfus 2009;
7:132–150.