reliability of broselow tape – length based emergency

116
RELIABILITY OF BROSELOW TAPE LENGTH BASED EMERGENCY WEIGHT ESTIMATION IN PAEDIATRIC POPULATION OF SOUTHERN TAMILNADU DISSERTATION SUBMITTED FOR THE DEGREE OF M.D BRANCH VII (PAEDIATRIC MEDICINE) APRIL 2018 THE TAMILNADU DR. M.G.R MEDICAL UNIVERSITY CHENNAI, TAMIL NADU

Transcript of reliability of broselow tape – length based emergency

RELIABILITY OF BROSELOW TAPE – LENGTH BASED EMERGENCY

WEIGHT ESTIMATION IN PAEDIATRIC POPULATION OF SOUTHERN

TAMILNADU

DISSERTATION SUBMITTED FOR THE DEGREE OF

M.D BRANCH VII

(PAEDIATRIC MEDICINE)

APRIL 2018

THE TAMILNADU DR. M.G.R MEDICAL UNIVERSITY

CHENNAI, TAMIL NADU

CERTIFICATE

This is to certify that the dissertation entitled “RELIABILITY OF BROSELOW

TAPE – LENGTH BASED EMERGENCY WEIGHT ESTIMATION IN

PAEDIATRIC POPULATION OF SOUTHERN TAMILNADU” is the

bonafide work of Dr. D. SANJEEV KUMAR in partial fulfilment of the

university regulations of the Tamil Nadu Dr. M.G.R Medical University, Chennai,

for M.D Degree Branch VII – PAEDIATRIC MEDICINE examination to be held

in April 2017.

Dr. D. MARUTHUPANDIAN M.S., FICS, FAIS

Dean, Madurai Medical College,

Government Rajaji Hospital,

Madurai – 625020

BONAFIDE CERTIFICATE

This is to certify that the dissertation entitled “RELIABILITY OF BROSELOW

TAPE – LENGTH BASED EMERGENCY WEIGHT ESTIMATION IN

PAEDIATRIC POPULATION OF SOUTHERN TAMILNADU” submitted by

Dr. D. SANJEEV KUMAR to the faculty of Pediatrics, The Tamil Nadu Dr.

M.G.R Medical University, Chennai in partial fulfillment of the requirement for

the award of M.D Degree Branch VII (PAEDIATRIC MEDICINE) is a bonafide

research work carried out by him under our direct supervision and guidance.

Dr. M.S.RAJARAJESWARAN MD DCH

Professor of Paediatrics

Institute of Child Health

& Research Centre,

Madurai Medical College,

Madurai

Dr. K. MATHIARASAN MD DCH

Director & Professor of Paediatrics

Institute of Child Health

& Research Centre,

Madurai Medical College,

Madurai

DECLARATION

I, Dr. D. SANJEEV KUMAR, solemnly declare that the dissertation titled

“RELIABILITY OF BROSELOW TAPE – LENGTH BASED

EMERGENCY WEIGHT ESTIMATION IN PAEDIATRIC POPULATION

OF SOUTHERN TAMILNADU” has been conducted by me at Institute of Child

Health and Research Centre, Madurai under the guidance and supervision of Prof.

Dr. M. S. RAJARAJESHWARAN M.D.,DCH..

This is submitted in part of fulfillment of the regulations for the award of

M.D Degree Branch VII (Paediatric Medicine) for the April 2017 examination to

be held under The Tamil Nadu Dr. M.G.R Medical University, Chennai. This has

not been submitted previously by me for any Degree or Diploma from any other

University.

Place : Madurai Dr. D. SANJEEV KUMAR

Date :

CONTENTS

Sl. No Title Page no

1. Introduction 1

2. Review of Literature 5

3. Aims and Objectives 28

4. Materials and Methods 29

5. Observation and Results 36

5. Discussion 61

6. Conclusion 67

7. Strength of this Study 69

8. Limitations 70

9. Recommendations 71

Annexures :

Bibliography

Proforma

Abbreviations

Master Chart

Ethical Clearance

Plagiarism Certificate.

ACKNOWLEDGEMENT

First, I would like to thank the almighty for giving me this

opportunity. My sincere thanks to Prof. Dr. D. MARUTHUPANDIAN, Dean,

Government Rajaji Hospital and Madurai Medical College for permitting me to do

this study and utilize the institutional facilities.

I express my sincere thanks and gratitude to Prof. Dr. K. Mathiarasan,

Professor and Director, Institute of Child Health & Research Centre, Madurai, for

his able supervision, encouragement, valuable suggestions and support for this

study.

I am greatly indebted to my teacher, Prof. Dr. M. S. Rajarajeswaran who

guided me throughout my study. I am also greatly thankful for his able supervision,

critical review, constant encouragement and full support rendered in every aspect

of this study..

I would extend my sincere thanks to Prof. Dr. S. Balasankar, Prof. Dr. M.

Kulandaivel, Prof. Dr. S. Shanmugasundaram and Prof. Dr. Nandhini, Prof.

Dr. M. Balasubramaniam, Prof. Dr. N. Muthukumaran for their valuable

advice and encouragement at every stage of this study.

I wish to express my sincere thanks to my Assistant Professors of Pediatrics,

Dr. P. Murugalatha, Dr. P. Ramasubramaniam, Dr. P. Kannan, Dr. K.

Ramya for their constant guidance, encouragement and support throughout my

study. I also extend my thanks to Dr. D. Rajkumar, Dr. E. Sivakumar, Dr. P. Guna,

Dr. J. Balasubramanian, Dr. S. Murugesalakshmanan, Dr. Vanitha, Dr. R. Suresh

for their guidance, supervision, valuable suggestions and support throughout this

study.

I thank the Institutional Ethical Committee for granting me permission to

conduct the study. I also express my gratitude to all my fellow Postgraduates for

their kind cooperation in carrying out this study and for their critical analysis.

Last but not the least, I submit my heartfelt thanks to the children and their

parents for extending full co –operation to complete my study successfully.

1

INTRODUCTION

Resuscitation of a critically ill child in the Emergency Department (ED)

adds an emotional stress to the treating physician. Calculating, prescribing,

preparing, and delivering accurate drug doses to children in the ED can be

challenging at the best of times and consequently medication errors in children

have been shown to be very common. Reliable paediatric weight estimation is the

keystone in child resuscitation in the ED. Knowledge on estimated body weight

is crucial in drug dosage calculation, determination of tube size, intravenous drug

administration, defibrillation and other life-saving measures(1). A series of steps

(as shown in figure 1) must be followed to avoid drug dose-related errors and

there is potential for medication error at any stage of the drug identification, dose

calculation, drug dilution or drug/dilution delivery. Flawed weight estimation

may lead on to negative effects on the child during resuscitation such as non-

responsiveness or increased adverse effects to the intervention(2).

Therefore, it is evident that accurate estimate of weight is imperative in

drug dose calculation system to

Ensure the efficacy of the drug administered

Prevent the plausible complications and side-effects of overdosing the

patient

2

Critically ill child in the resuscitation room

Identification of correct medication

Determination of correct base dose (mg/kg)

Correct calculation of total dose WEIGHT

REQUIRED

Accurate verbal or written prescription

Correct interpretation of prescription

Accurate determination of drug dilution

Accurate performance of drug dilution

Accurate calculation of ml volume to deliver

Accurate delivery of correct volume

Figure 1: The sequential steps required for the correct calculation, dilution,

preparation and delivery of medications to children during emergency

medical management or resuscitation.

3

The ideal method for weight estimation is to weigh the child on a scale,

which is often inappropriate in emergency situations as this requires fixed

equipment and adequate time(3). In developing countries like India and in

emergencies, limited time or resources makes it difficult for actual weight

measurement(4,5).

Multiple methods of weight estimation on the basis of age alone or on a

combination of age and height have been reported in the literature(6-12). Age based

weight estimation was preferred because of the ease of calculation, reliability and

rapidity. Recently the accuracy of these methods have been challenged(6,13,14) as

these methods gives rise to a significant deviation in weight estimation(12).

Most methods to estimate weight from age are based on western pediatric

populations(6,7). In India, the age-derived weight methods using the Nelson

formula or APLS in pediatrics is often used(8,18). There are no standardized

methods in use for rapid weight estimation in children admitted for acute pediatric

emergencies in trauma centers in Madurai, India. Physicians currently make an

educated guess from experience and/or ask parents for the last known weight.

The Broselow Tape (Broselow Paediatric Emergency Tape, 2002 Edition,

Drs James Broselow and Robert Luten, Armstrong Medical Industries, 575

Knightbridge Parkway, Lincolnshire, IL 60069, USA) was developed using

height/weight correlations from Western data. This validated tape estimates

weight (hence intervention doses/sizes) of the supine child based on length(15,16).

4

The Broselow tape has been conclusively shown to improve accuracy in weight

prediction and eliminate the need for memorization and calculation in such

situations(17). This study attempts to investigate the accuracy and clinical utility

of the Broselow tape in paediatric population of Madurai, India.

The rationale behind my study is that, I found the Broselow-Luten tape is

underutilized in our setup as the medical personnel or the staff preferred age based

mathematical formulae.

5

REVIEW OF LITERATURE

Management of a critically ill child in the Emergency Department (ED)

requires greater confidence for optimum care as they are a diverse population in

weight, size, shape and developmental level(13). Obtaining an accurate weight of

a child allows for optimum care of the child. Dosage errors from wrong weight

calculation contributes to most adverse drug events(17).

Measuring the weight of a child requires time and a systematic and uniform

approach. This literature review will delineate the common methods of weight

estimation used in healthcare system and analyse current study in relation to

these. Different methods of weight estimation (as shown in table 1) were followed

and are broadly classified into

1. Age based

2. Length based

3. Visual estimation

4. Other methods

From the table 1, it is evident that there are numerous weight estimation

methods. Nonetheless, methods which measure the precision of weight

estimation are not consistent across studies. Authors report accuracy based on

selection of clinically relevant cut-off; most frequently, weight estimates within

10% of a child’s measured weight(22,25,26). Some authors use mean percentage

error, mean percentage difference or actual error in kg to report statistical

significance(6,12,13).

6

TABLE 1: METHODS OF WEIGHT ESTIMATION

AGE BASED LENGTH

BASED

VISUAL

ESTIMATION

OTHERS

1. Advanced

Paediatric Life

Support (APLS)

2. Australian / New

Zealand

Resuscitation

Council

(ARC/NZRC)

3. Argall

4. Best Guess

Formula

5. Chinese Weight

Assessment Rule

6. Leffler

7. Luscombe&Owens

8. Nelson

9. Park Formula

10. Shann

11. Theron

12. Devised

Weight

Estimation

Method

(DWEM)

13. Broselow-

Luten Tape

14. Kloek Tape

15. Lo Tape

16. Malawi Tape

17. Oakley Tables

18. PAWPER

Tape

19. PREM Tape

20. Sandell Tape

21. Traub-Johnson

22. Traub-Kichen

23. Mercy Tape

24. Physician

25. Ambulance

26. Parent

27. Nurse

28. Clothing Label

Size

29. Haftel Hanging

Leg

30. Arm

Circumference

7

History of weight estimation methods in children

In the early 1980s, age based weight estimation methods and educated

guesses by the treating physician were commonly used for calculating weight in

seriously ill children in Emergency Department(7).

The earliest record of age based formula of weight estimation available in

literature was given by Alexander Weech in 1956(31) with a series of equations to

predict height and weight in children in metric and imperial units. Weech

formulas incorporated the one which later become as the APLS formula.

The likely time of introduction of APLS formula is sometime from 1952

to 1956, derived from post war children in the UK. Traub and Johnson(32) and

Traub and Kichen(33) formulas were developed in early 1980s.

8

FL

OW

CH

AR

T 2

: T

IME

LIN

E F

OR

WE

IGH

T E

ST

IMA

TIO

N M

ET

HO

DS

9

TABLE 2: HISTORY IN A NUTSHELL

YEAR

AUTHOR

NAME

METHOD

1986

Jeffrey Garland(34)

Devised Weight-

Estimating

Method (DWEM)

Only technique

which includes

body habitus in a

weight estimation

technique

1988

Peter Oakley(35)

Reference table of

Weight-for-length

for children.

1988

Deborah Lubitz(7)

First validation

study on the

Broselow

Paediatric

Emergency Tape

Early 1980s

James Broselow(19)

Broselow tape

Length based

Weight

estimation

method, based on

50th centile of the

1979 NCHS

weight-for-length

growth charts

1990

Anthony Haftel

Hanging leg-

weight technique

1997

Leffler(36)

Age based weight

estimation

method

2000

Shann(37) Age based weight

estimation

method

2003

Argall(3)

Age based weight

estimation

method

2004

Nelson(18)

Age based weight

estimation

method

2007

Luscombe(12)

Age based weight

estimation

method

10

2007

Tinning(10)

Best Guess

Age based weight

estimation

method

OTHER TAPE BASED WEIGHT ESTIMATION METHODS

1995 Lo tape / rod

modification(37)

Developed in

Hong Kong

1999 Elizabeth Molyneux Blantyre tape(38) Malawi

2000 Walter Kloeck Kloeck tape(39) South Africa

2004 Julian Sandell Sandell tape(40) UK

2005 Joe Brierly Paediatric

Resuscitation

Emergency

Management

(PREM) tape(41)

2008 Michael David John

Wells

Paediatric

Advanced Weight

Prediction in the

Emergency Room

(PAWPER) tape

11

WEIGHT ESTIMATION METHODS

1. LENGTH BASED

(i) Broselow-Luten tape: James Broselow, a North Carolina–based

emergency physician, developed a simple tool to improve weight estimation

using established height-weight correlations in the early 1980s(19). The Broselow

tape is a colour-coded tape which gives a weight estimate based on child’s length

from head to heel. The Broselow tape is a full-scale depiction of the NCHS 2000

weight-for-length growth chart, with the weight divisions of the tape representing

the 50th centile. The tape is calibrated in kilograms rather than any measurement

of length, and information relating to drug dosaging or equipment sizing may be

found within each weight division This Tape has been validated in western

population and has shown to decrease errors in paediatric weight estimation

particularly in emergency department.

Based on the literature available, it is said that the most common paediatric

drug errors are dosage errors and are often based on flawed weight estimates(20).

Previous versions of Broselow tape gave a weight estimate, but further

calculations were required to obtain the medication dose. Newer versions has

resolved these problems and the printed versions of the Broselow tape contains

weight estimated by length and resuscitation drug doses and equipment size

included to decrease the risk of error.

The Broselow-Luten Tape was first validated in the USA in 1988 in a study

by Lubitz, et al., set in Emergency Departments and Outpatients Clinics. They

12

concluded that the SD remains small for all age groups (<3.67), giving it

precision, nonetheless, when assessing the total population actual error spans 15

kg, which according to Mackway-Jones et al could cause a significant difference

in drug dosage or equipment size during resuscitation. Lubitz et al., showed in

their study that Broselow-Luten tape was highly precise in comparison with other

methods which were in use during that time period. However, most studies in

Table 3 show that with increasing age or weight, the accuracy of the Broselow-

Luten Tape decreases. One exception to this finding is a South African study in

2011 where the Broselow tape estimated 65.38% of 6-10 years children as within

10% of measured weight which is of more precision then the younger age group

which is also shown in table 3.

There have been several iterations of the Broselow-Luten Tape with

updates to weight and drug information. The most recent of which is the 2011

version, which is used in this study. This recent version of the Broselow tape has

its own limitations such as the data used to create this was from North American

children making it difficult for use in other parts of the world(21). However, many

studies worldwide have found length-based methods such as the Broselow-Luten

tape, the most accurate of the weight estimation methods tested(6.13,22,23,24).

Main limitation of any length based weight estimation method is that,

during measurement of weight, child needs to be either in a standing posture or

laying flat. This is practically difficult(22) in some emergency resuscitation

situations, for example in cases of severe asthma, laying flat is contraindicated

13

making length based methods of estimation ineffective. Since Broselow tape has

been the most widely used method of weight estimation during resuscitation for

the last few decades, it has also been the object of the most critical review in

literature.

Some of the concerns relevant to the use of the Broselow tape in Madurai, India

include:

It has not been formally validated in a South Indian population.

It has been shown to underestimate weight in overweight children(42) and

overestimate weight in underweight children(43). As a result it has proved to be

inaccurate in populations where a significant proportion of children are above

average weight-for-length.

Since the number of overweight children is increasing in many countries

worldwide, Including India, the potential for an escalating underestimation of

weight is concerning. There is also, on the other hand, a sizeable population of

undernourished children in India who are at risk of having an overestimation of

weight. Our population is thus at risk at both ends of the spectrum for inaccurate

weight estimation. Hence, we decided to validate this tape in our population for

further use during resuscitation in Emergency Department.

14

TABLE 3: Summary of studies assessing the accuracy of the Broselow-

Luten Tape using percent of estimates within 10% of measured weight

Authors N Age/Weight

group

Within 10% LOA Country Setting

Lubitz et

al (1988)

395

449

93

937

3.5 to 10 kg

10 to 25 kg

> 25kg

Total

55.9%

65.0%

49.5%

59.7%

SD 0.95

SD 1.62

SD 3.67

AE -5.7 to

15.7 kg

USA Prospective,

observational

multicentre

study in ED &

OPD

Nieman,

Manacci,

Super,

Mancuso

and Fallon

(20006) –

2249

1403

1224

2937

Infant

Toddlers

Preschool

School Aged

59.5% /

55.0%

64.3% /

60.0%

61.4% /

59.4%

57.6% /

51.2%

USA Prospective,

observational

study set in

School &

Paediatric

Clinic. Tape

Version 1998 /

2002A

Abdel-

Rahman et

al. (2013a)

976 2 – 14 years 55.3% USA Prospective,

multicentre,

observational,

children from

Hospital

Milne,

Yasin,

Knight,

Noel,

Lubell and

Filler

(2012)

6361 <10 years 56.3% -7.71 –

10.95 kg

Canada Prospective,

observational

study in Urban

& Rural Health

Centre /

School

Geduld et

al. (2011)

2832 1-4 years

6-10 years

63.46%

65.38%

±0.65

±0.90

South

Africa

Prospective via

database at

hospital

Trakulsric

hai et al.

(2012)

300

(ED)

295

(OPD)

< 10 kg

10 – 25 kg

25 – 40 kg

> 40 kg

60.36%

59.68%

43.20%

N/A

-2.68 - 6.31

-3.91 - 4.48

-13.22- 7.26

Thailand Prospective

observational

study in OPD

& ED

15

FIGURE 2: BROSELOW PAEDIATRIC EMERGENCY TAPE

16

(ii) Oakley Chart: developed in 1988 using a reference chart from which

weight could be read after the child’s age was determined or height or length was

measured. It was based on the same principle as that of growth chart method. This

method was used in some parts of the world, but the chart can be found only in

the original publication(44). Major limitations of the chart was, it is derived from

old growth charts making it a time consuming one to use. Also literature shows

that this method has been found to significantly underestimate the weight of

children, most specifically in children >25 kg(6).

(iii) DWEM Table: developed in 1986, which is produced with a table of

weights for children with slim, average and heavy frames at 50mm intervals of

length based on th 5th (slim), 50th (average) and 95th (heavy) 1979 NCHS weight-

for-length growth chart centiles. This method involves measuring child’s length,

assessing their frame size and habitus and then reading the appropriate weight off

the table. Because of its complex nature in the measurement of weight, it has not

seen much clinical use since its inception.

(iv) Traub – Johnson formula:

Weight (kg) = 2.05 x e0.02X

Used for children aged 1 to 18 years. X = height or length in cm.

(v) Traub – Kichen formula:

weight (kg) = 2.396 x 1.0188X

Used in children over 74cm and aged 1-17 years. X = height or length in cm.

17

Both these equations are formidable to the treating emergency physician

and requires use of a scientific calculator which makes the use of these formulas

vulnerable to failure and to errors which might not be readily detectable.

(vi) Blantyre tape: developed in Malawi in 1999 as an alternative to

Broselow tape as the authors felt that Broselow tape overestimated weight in their

community and also owing to the fact that they did not have the same drugs

available that were represented on the Broselow tape(38). This tape is a hand drawn

tape with divisions at 50mm intervals corresponding to a weight reflecting 85%

to 90% of the 50th centile of the NCHS 2000 weight-for-length chart.

The validation study of the tape had some methodological shortcomings in

the design and analysis, particularly that a large margin of error in weight

estimation (20%) was considered acceptable. The tape is still in use in the Queen

Elizabeth Central Hospital in Blantyre but has not been tested in successive

studies.

(vii) Sandell tape: This tape is much similar to the Broselow tape in that

it embraces information regarding drug dose, emergency equipment sizes related

to length. It is based on the 1996 UK growth charts and was designed primarily

for the prehospital environment. This tape was used by ambulance services in

UK. There are no studies assessing it utility or accuracy.

(viii) PREM tape: The Paediatric Resuscitation Emergency Management

(PREM) tape was developed in United Kingdom, the basis of which is the

UK1996 growth charts. This tape is disposable, exclusively used to estimate

18

weight with an accompanying booklet which helps to calculate drug dose and

equipment sizes. No studies are available in literature validating its precision.

(ix) Kloeck tape: developed in South Africa by Walter Kloeck. It is an

inexpensive, laminated paper tape. On one side of the tape, regular tape measure

is printed and the other side has the same colour zones as that of Broselow tape.

No validation studies available in the literature.

(x) PAWPER tape: developed in South Africa. This tape includes both a

measure of length and of body habitus in the methodology. With this tape, if a

length based weight has been estimated, this weight can be altered according to

the child’s body habitus. The weight can be acclimatized up or down or left

unmodified based on whether the child is thought to be overweight, underweight

or of average weight, respectively. The weights for children with above or below

average weight-for-length can be found on the tape itself, within each weight

division.

2. AGE BASED WEIGHT ESTIMATION METHODS

(i) Advanced Paediatric Life Support (APLS) weight estimation

formula: This method is used worldwide and inculcated during Advanced

Paediatric Life Support (APLS) courses worldwide(8). The first reference to the

APLS course was in the Journal of Paediatrics in 1982 as a scheduled course,

content was not disclosed and therefore, it can only be assumed that a weight

estimation formula was included in this course ("Scheduled meetings," 1982).

19

The most well known formula is that taught on the Advanced Paediatric Life

Support (APLS) course(27):

Weight (kg) = (age in years + 4) X 2 (or)

Weight (kg) = (2 X age in years) + 8

This formula was derived from National Centre for Health Statistics

(NCHS) population data from 1977 in the United States(28). However, subsequent

studies from the USA and the UK have suggested that the APLS formula

underestimates children’s weight by up to 20% in developed countries(6,29). In the

United Kingdom, the APLS formula has been updated to reflect the increasing

weight of children aged 6 to 12 years (Resuscitation Council United Kingdom,

2012):

1 - 12 months = (0.5 x age months) + 4

1 - 5 years = (2 x age years) + 8

6 - 12 years = (3 x age years) + 7

A study of New Zealand Maori and Pacific children by Theron et al. (2005)

showed that these populations were considerably heavier than others in Auckland,

therefore, the APLS formula consistently underestimated the weight of these

children. The variability in weight-for-age is significantly higher than that for

weight-for-length and therefore none of these age-based formulas have ever been

shown to perform better than length based systems(6,43). Validation studies

performed to assess the accuracy of age based weight estimation methods is given

in table 4 & 5.

20

TABLE 4: VALIDATION STUDIES PERFORMED TO ASSESS THE

ACCURACY OF APLS FORMULA

FORMULA

POPULATION

ERROR

APLS

UK

Australia

Pacific Islands

Australia

Australia

Malawi

India

<20% within 10% of actual weight

34% within 10% of actual weight

95%CI -7.7kg to 13.4kg

MPE -15%

MPE -17%

MPE 10%

Mean overestimate 2kg

MPE – Mean Percentage Error

21

Most experts advocate that age-based formulas should only be used if no

better weight-estimation system is available(45). One of the advantages of age-

based formula methods is that they are independent of any equipment and so

constitute a highly resilient weight-estimation technique, as long as the child’s

correct age is known, the formula is remembered correctly and the arithmetic

performed accurately. The memorisation of age-based formulas is encouraged on

advanced life support courses, but memory is capricious in emergencies and leads

to mistakes(46).

(ii) Nelson’s formula:

Children aged 3 – 12 months = [age in months + 9] / 2

Children aged 1 – 6 years = 2 x [age in years + 4]

Children aged 7 – 12 years = {[age in years x 7] – 5} / 2

This method has only been evaluated in only one published study from a

population of Indian children which showed poor overall accuracy (a mean error

of 5.2 ± 4.3kg) and particularly in children under 12 months and over 7 years of

age(5).

(iii) Argall’s formula: developed in UK. Validation studies included only

a sample of children in UK and studies outside UK yielded poor results.

Weight (kg) = 3 x (completed years of age + 2)

This formula is used in children from 1 – 10 years of age.

22

(iv) Luscombe’s formula: This formula was also developed in a

population of children from UK. This formula has not been validated outside of

the original study population.

Weight (kg) = (3 x completed years of age) + 7

This formula is used in children aged 1 – 10 years.

(v) Shann’s formula:

Weight (kg) = (age in years x 2) + 9

This formula is for children from 1 – 9 years of age.

Weight (kg) = (age in years x 3)

This formula is for children aged >9 years.

The Shann formula has been shown to perform better than other age-based

formulas in Maori and Pacific Island children(13). The analysis in this study was

methodologically incomplete and therefore the conclusions might not be valid.

This formula has not found universal acceptance.

(vi) Leffler’s formula:

Weight (kg) = (age in months + 8) / 2

This formula is used in children <1 year of age.

Weight (kg) = (age in years x 2) + 10

This formula is used in children aged 1 – 10 years of age.

The Leffler formula has been evaluated only in a single study since its

publication in 1997. In this study it was shown to perform no better than the Shann

23

and APLS formulas, demonstrating a significant error in both bias and

precision(13).

(vii) Best Guess (Tinning’s) formula:

Weight (kg) = (age in months + 9) / 2

This formula is used for children ≤12 months of age.

Weight (kg) = (age in years x 2) + 10

This formula is used for children aged 1 – 5 years.

Weight (kg) = age in years x 4

This formula is used for children aged 6 – 14 years.

The “Best Guess” formulas were developed in Australia. The validation

studies published failed to show acceptable accuracy of these formulas(47,48).

24

TABLE 5: VALIDATION STUDIES PERFORMED TO ASSESS THE

ACCURACY OF OTHER AGE BASED FORMULAS

FORMULA

POPULATION

ERROR

TINNING

Australia

Australia

Australia

42% within 10% of actual weight

MPE 8%

<50% within 10% of actual weight

LUSCOMBE

Malawi MPE 17%

SHANN

Pacific Islands 95%CI -8.7kg to 12.3kg

ARGALL

Australia

Australia

37% within 10% of actual weight

37% within 10% of actual weight

TRAUB-JOHNSON

Australia MPE 10%

TRAUB-KICHEN

Australia MPE 12%

NELSON

India Mean Overestimate 2kg

LEFFLER

Pacific Islands 95%CI -7.3kg to 11.3kg

MPE – Mean Percentage Error

25

3. ESTIMATION OF WEIGHT BY THE PARENT/S

The applicability of parental estimates of their child’s weight is dependent on

the parent being present and accessible to healthcare personnel at the time of the

child’s presentation in the ED. The precision of weight is decided by knowing if

the parent is the regular caregiver of the child or not. Usually parents recall the

recently recorded weight, which is assumed to be more accurate than any

estimation. However, reports are conflicting about the ability of parents to

accurately estimate the weight. But, in general, parents can estimate the weight

of their children better than healthcare workers and with sufficient accuracy for

acceptable resuscitation drug dose calculation(4,35,49,50).

4. ESTIMATION OF WEIGHT BY THE HEALTHCARE PROVIDER

The efficiency of Emergency Physician, Professional nurses and supporting

paramedics to estimate the weight of children precisely in emergency situations

remains limited(51-53). Previous literature shows that children who fall within 10%

of their actual body weight is 25 – 50%. This fallacy is lesser in infants compared

to older children(53). Usually the health care provider estimates weight of children

by either one of the following methods. If age of child is known, they use age

based weight estimation formulas (like APLS etc.). If age of the child is not

known, a pure guess based on the three dimensional conformation of the child is

made which is less commonly used and the least accurate method.

26

5. GROWTH CHART METHOD

This method is generally not convenient in an emergency setting and it used

weight for age or weight for length growth charts for estimating weight from

child’s age or measured length. It is relatively a time-consuming method

compared to other methods of weight estimation. Using a growth chart needs

expertise and acquaintance with those charts. Minor differences between various

growth charts exist, however, 50th centiles are almost same and the choice of

growth chart to be used is not critical(54,55)

OTHER METHODS OF WEIGHT ESTIMATION

(i) The hanging leg-weight technique: was developed in 1990 by

Anthony Haftel and his fellow emergency physicians from Los Angeles

Children’s Hospital(56). The authors made a measurement of 100 anaesthetised

children with a spring scale, both their legs suspended in a sling around the heels.

The sling was raised to 450 from horizontal, at which point hanging leg weight

was recorded. This was then compared with the total body weight and found it to

predict body weight precisely, especially in children >10kg body weight.

Weight (kg) = (5.176 x Hanging leg weight in kg) + 3.487

(ii) The infant foot-length technique: is an indigenous technique

developed in India. A group of paediatricians under the leadership of Sandeep

Bavdekar met at KEM Hospital, Mumbai in 2006 and developed this formula to

estimate the weight of infants <2 years of age by measuring the foot-length(60).

27

Weight (kg) = -5.15 + (foot length in cm x 1.35)

The study showed that foot length was compatible with body weight and

can be used in infant resuscitations. The authors then created a table of foot length

(in 1 mm increments from 45mm to 80mm) with the corresponding weight

(0.94kg to 5.67kg) and doses of resuscitation drugs. It was found that the study

had methodological weakness in the analysis part with respect to the accuracy

and precision of the technique, however, the technique appears to be promising.

No furher studies were done in India or elsewhere to support the technique.

(iii) Carroll technique: described by Carroll and colleagues in 2001(61), to

predict weight from the coalition of an age based weight estimation formula; shoe

size and mid arm circumference. This formula is too difficult to use in emergency

department.

28

AIMS AND OBJECTIVES

1. To determine the degree to which the length based weight estimation tape

(Broselow tape) predicts measured weight in children of Madurai.

2. To validate the length based weight estimation tape (Broselow tape) for

use in our clinical setting.

29

MATERIALS AND METHODS

The design was a prospective observational hospital based study, inclusive

of children aged less than 12 years attending the Outpatient department, Institute

of Child Health & Research Centre (ICH&RC), Government Rajaji Hospital,

Madurai. The study was conducted over a period of one year (April 2016 – March

2017). This tape measures between 46 cm – 143 cm and children with length of

this range only were included apart from other exclusion criteria. This

measurement has got an added advantage that the growth spurt which occurs

during adolescent age group will be nullified, avoiding bias.

SAMPLE SZE ESTIMATION

N=size per group; d=; δ0= the real mean difference between two group; S2=

Polled standard deviation of both comparison groups. ἀ= precision (0.05) and =

90%.

To detect a 5% difference between the Broselow-predicted weight and the

Actual weight a minimum sample size of 204 children was required in each

subgroup to produce a power of >0.9, with a minimum aggregate of 408 children.

30

Exclusion criteria of the study were, children who

require emergency resuscitation

are severely dehydrated

have joint contractures

Institutional ethical committee approval was obtained.

After getting an informed parental consent, data were collected which

included demographic information such as gender, age in years. The Broselow

tape weights were determined, according to the tape directions. Child was made

to lie in supine position with head kept at neutral position with no pillow in place.

One end of the tape mentioned “MEASURE FROM THIS END” which was kept

at the vertex and then the tape was stretched to the child’s heel by straightening

the hip and knee joints while the ankle was flexed at 90°. The heel markings were

made on the tape and child’s weight was noted from the tape at that point. Height

or Length is not reported by the Broselow tape and is not a variable in our study.

Then the child was weighed using a child weighing scale or baby weighing scale.

Child was either undressed or minimally dressed. These weighing scales were

calibrated daily.

Performance of Broselow tape was compared with the actual weight ie.,

reference weight in order to validate the tape. Other factors considered were the

difference in tape predicted weight and reference weight which is considered as

a measure of estimation bias, the standard deviation of the difference in weight

31

PICTURE 1: WEIGHING MACHINE USED IN OUR STUDY

32

which is considered as a measure of estimation of precision, measure of tape

accuracy.

STATISTICAL ANALYSIS

The data collected were entered in Microsoft excel sheet 2010. Results

were analyzed using the SPSS version 20 (IBM corporation, New York, U.S.A).

Reporting of the data was done using mean with standard deviation and 95%

confidence interval for parametric and interval data; median with interquartile

range for ordinal and nonparametric interval data.

Agreement between the weight predicted by Broselow tape and the

reference weight was determined by Spearman correlation analysis.

The difference between estimated and measured weight was calculated

using, difference (kg) = predicted weight – measured weight. These difference

were considered in Bland-Altman graphs, which uses a reformed technique to

plot the difference against the measured weight, rather than the mean of the

estimated and measured weights. Bland-Altman technique is a standardised

analysis used to compare two measurement or analysis methodologies to provide

a measure of bias and precision. With this technique, accuracy or bias is

determined from the mean of the difference between the two values, while

precision is determined from the standard deviation of those differences. This was

used in literature that have validated Broselow tape previously.

The percentage difference (MPE) between weights predicted by the

Broselow tape and the actual weight was calculated using

33

Error % = 100 X [predicted weight – reference weight] / reference weight

These percentage errors were compared and analysed using the t-test for

dependent samples, both for the entire sample as well as the different genders,

weight categories.

The percentage error values of each of the measurement techniques were

categorised:

Within 5% of actual weight.

Within 5 to 10%.

10 to 15%.

15 to 20%.

>20% error.

P value of <0.05 was considered statistically significant.

34

PICTURE 2: MEASURING WEIGHT BY BROSELOW TAPE

36

OBSERVATION AND RESULTS

During the study period, 1099 children were included in our study. 211

children were grouped under 10kg, 543 children were included in 10 – 20kg

group, 250 children in 20 – 30kg group and 95 children in >30kg group as shown

in table 6.

TABLE 6: POPULATION CHARACTERISTICS – NUMBER OF CASES

GROUP NUMBER OF CASES (N)

<10kg 211 (19.2%)

10 – 20kg 543 (49.4%)

20 – 30kg 250 (22.8%)

>30kg 95 (8.6%)

TOTAL 1099 (100%)

The study enrolled 748 (68.1%) male children and 351 (31.9%) female

children. The overall mean ± standard deviation for age was 4.6 ± 3.1 years

(median age 4.0 years). The mean ± standard deviation for measured weight

was 17.5 ± 8.3kg (median 15kg). The mean ± standard deviation for

Broselow tape weight was 19.2 ± 9.8kg. Demographic characteristics

are summarised in table 7.

37

TABLE 7: DEMOGRAPHIC CHARACTERISTICS

WEIGHT GROUPS

<10kg

10 – 20kg

20 – 30kg

>30kg

TOTAL

Number

(N)

211

543

250

95

1099

Age (years)

mean ± SD

1.0 ± 0.5

3.65 ± 1.55

7.6 ± 1.59

10.29 ± 1.3

4.6 ± 3.1

Male 141

(66.8%)

286

(71.1%)

167

(66.8%)

54

(56.8%)

768

(68.1%)

Female 70

(33.2%)

157

(28.9%)

83

(33.2%)

41

(43.2%)

351

(31.9%)

Actual

weight (kg)

Mean ± SD

7.7 ± 1.85

14.7 ± 2.9

25.8 ± 3.0

34.0 ± 3.1

17.5 ± 8.3

Broselow

weight (kg)

mean ± SD

8.6 ± 2.15

15.4 ± 3.7

28.8 ± 5.1

38.3 ± 3.5

19.2 ± 9.8

38

FIGURE SHOWING DISTRIBUTION OF CASES

0

100

200

300

400

500

600

<10 Kg 10 - 20 Kg 20 - 30 Kg >30 Kg

NUMBER OF CASES

NUMBER OF CASES

39

The correlation between actual weight and tape weight was excellent, with

the r2 values for the Broselow tape being 0.97. Correlation was generally best in

lesser weight groups ie., infants and younger children when compared with older

children. Table 8 shows the correlation between actual measured weight and the

Broselow tape weight. Figure 3-7 shows correlation between actual and broselow

weight.

TABLE 8

CORRELATION BETWEEN ACTUAL AND BROSELOW WEIGHT

Measured Weight Groups

Overall <10kg 10 – 20kg 20 – 30kg >30kg

Correlation

coefficient

0.97 0.86 0.89 0.78 0.44

Significance

level

p<0.001 p<0.001 p<0.001 p<0.001 p<0.001

40

FIGURE 3: CORRELATION BETWEEN ACTUAL AND BROSELOW

WEIGHT (OVERALL STUDY)

FIGURE 3: CORRELATION BETWEEN ACTUAL AND BROSELOW WEIGHT

Bro

selow

weigh

t

41

FIGURE 4: CORRELATION BETWEEN ACTUAL AND BROSELOW

WEIGHT (<10Kg)

FIGURE 4: CORRELATION BETWEEN ACTUAL AND BROSELOW

WEIGHT (<10Kg)

Bro

selow

weigh

t

42

FIGURE 5: CORRELATION BETWEEN ACTUAL AND BROSELOW

WEIGHT (10-20Kg)

FIGURE 5: CORRELATION BETWEEN ACTUAL AND BROSELOW WEIGHT

(10-20Kg)

43

FIGURE 6: CORRELATION BETWEEN ACTUAL AND BROSELOW

WEIGHT (20-30Kg)

FIGURE 6: CORRELATION BETWEEN ACTUAL AND BROSELOW WEIGHT

(20-30Kg)

Bro

selo

w w

eigh

t

44

FIGURE 7: CORRELATION BETWEEN ACTUAL AND BROSELOW

WEIGHT (>30Kg)

FIGURE 7: CORRELATION BETWEEN ACTUAL AND BROSELOW WEIGHT

(>30Kg)

Bro

selow

weigh

t

45

BLAND-ALTMAN ANALYSIS

The overall bias, as demonstrated by Bland-Altman methodology was an

underestimation of 1.6 kg for the Broselow tape. Precision of our study with 95%

confidence limits of the Bland-Altman methodology was -4.7kg to 3.5kg. Figure

8-12 shows Bland-Altman plots for weight estimation technique of the entire

population and of the sub-categories.

FIGURE 8: BLAND-ALTMAN PLOT SHOWING BIAS AND LIMITS OF

AGREEMENT FOR THE ENTIRE STUDY POPULATION

-4.7 -4.7

3.46 3.46

-1.61 -1.61

-20

-15

-10

-5

0

5

10

15

0 10 20 30 40 50 60

BR

OSE

LOW

WEI

GH

T-A

CTU

AL

WEI

GH

T (K

G)

Bland altman analysis LOA UOA BIAS

46

FIGURE 9: Bland-Altman plot illustrating the bias and limits of agreement

for the cohort of patients with a body weight of ≤10kg.

-0.92 -0.92

0.76 0.76

-2.6 -2.6

-3.5

-3

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

0 2 4 6 8 10 12 14 16

BR

OSE

LOW

WEI

GH

T-A

CTU

AL

WEI

GH

T (K

G)

Linear (UPPER LIMIT) Linear (LOWER LIMIT)

47

FIGURE 10: Bland-Altman plot illustrating the bias and limits of agreement

for the cohort of patients with a body weight of 10-20kg.

-0.75 -0.75

2.55 2.55

-4.03 -4.03

-14

-12

-10

-8

-6

-4

-2

0

2

4

0 5 10 15 20 25

Bro

selo

w w

eigh

t-A

ctu

al w

eigh

t (k

g)

Linear (BIAS) Linear (UPPER LIMIT) Linear (LOWER LIMIT)

48

FIGURE 11: Bland-Altman plot illustrating the bias and limits of agreement

for the cohort of patients with a body weight of 20-30kg.

-3.05 -3.05

2.85 2.85

-8.95 -8.95

-15

-10

-5

0

5

10

0 5 10 15 20 25 30 35

BR

OSE

LOW

WEI

GH

T -

AC

TUA

L W

EIG

HT

(KG

)

AXIS TITLE

Series1 Linear (BIAS) Linear (UPPER LIMIT) Linear (LOWER LIMIT)

49

FIGURE 12: Bland-Altman plot illustrating the bias and limits of agreement

for the cohort of patients with a body weight of >30kg.

-4.27 -4.27

4.2 4.2

-12.74 -12.74

-15

-10

-5

0

5

10

15

0 10 20 30 40 50 60

Series1 Linear (BIAS) Linear (UPPER LIMIT) Linear (LOWERLIMIT)

50

The other major indicator of bias, the mean percentage error (MPE),

showed an underestimation of 8.5%. And the precision for MPE, reflected by

Standard deviation was 11.49. The percentage of cases falling within each

percentage error category is shown in table 9.

TABLE 9: CATEGORISED PERCENTAGE ERROR DATA FOR

ENTIRE STUDY POPULATION

Percentage Error (%) Number of Cases (N) Percentage %

>-15 313 28.5

-10 to -15 200 18.2

-5 to -10 214 19.5

0 to -5 53 4.8

0 to 5 160 14.6

5 to 10 97 8.8

10 to 15 41 3.7

>15 21 1.9

TOTAL 1099 100

51

FIGURE 13: CATEGORISED PERCENTAGE ERROR DATA FOR

ENTIRE STUDY POPULATION

0

50

100

150

200

250

300

350

>-15 -10 to -15 -5 to -10 0 to -5 0 to 5 5 to 10 10 to 15 >15

NUMBER OF CASES

NUMBER OF CASES

52

TABLE 10: CATEGORISED PERCENTAGE ERROR DATA FOR <10kg

CHILDREN

Percentage Error (%) Number of Cases (N) Percentage %

>-15 82 38.9

-10 to -15 46 21.8

-5 to -10 30 14.2

0 to -5 8 3.8

0 to 5 30 14.2

5 to 10 6 2.8

10 to 15 8 3.8

>15 1 0.5

TOTAL 211 100

53

FIGURE 14: CATEGORISED PERCENTAGE ERROR DATA FOR <10kg

CHILDREN

NUMBER OFCASES

0

10

20

30

40

50

60

70

80

90

>-15 -10 TO -15 -5 TO -10 0 TO -5 0 TO 5 5 TO 10 10 TO 15 >15

NUMBER OF CASES

NUMBER OF CASES

54

TABLE 11: CATEGORISED PERCENTAGE ERROR DATA FOR

10 -20kg CHILDREN

Percentage Error (%) Number of Cases (N) Percentage %

>-15 80 14.7

-10 to -15 78 14.4

-5 to -10 140 25.8

0 to -5 28 5.2

0 to 5 99 18.2

5 to 10 83 15.3

10 to 15 22 4.1

>15 13 2.4

TOTAL 543 100

55

FIGURE 15: CATEGORISED PERCENTAGE ERROR DATA FOR

10 -20kg CHILDREN

0 20 40 60 80 100 120 140 160

>-15

-10 to -15

-5 to -10

0 to -5

0 to 5

5 to 10

10 to 15

>15

NUMBER OF CASES

NUMBER OF CASES

56

TABLE 12: CATEGORISED PERCENTAGE ERROR DATA FOR

20 -30kg CHILDREN

Percentage Error (%) Number of Cases (N) Percentage %

>-15 99 39.6

-10 to -15 53 21.2

-5 to -10 35 14

0 to -5 17 6.8

0 to 5

28 11.2

5 to 10

7 2.8

10 to 15

9 3.6

>15

2 0.8

TOTAL 250 100

57

FIGURE 16: CATEGORISED PERCENTAGE ERROR DATA FOR

20 -30kg CHILDREN

>-1540%

-10 to -1521%

-5 to -1014%

0 to -57%

0 to 511%

5 to 103%

10 to 153%

>151%

NUMBER OF CASES

>-15 -10 to -15 -5 to -10 0 to -5 0 to 5 5 to 10 10 to 15 >15

58

TABLE 13: CATEGORISED PERCENTAGE ERROR DATA FOR >30kg

CHILDREN

Percentage Error (%) Number of Cases (N) Percentage %

>-15 52 54.7

-10 to -15 23 24.2

-5 to -10 9 9.5

0 to -5 0 0.0

0 to 5

3 3.2

5 to 10

1 1.1

10 to 15

2 2.1

>15

5 5.3

TOTAL 95 100

59

FIGURE 17: CATEGORISED PERCENTAGE ERROR DATA FOR >30kg

CHILDREN

0 10 20 30 40 50 60

>-15

-10 to -15

-5 to -10

0 to -5

0 to 5

5 to 10

10 to 15

>15

NUMBER OF CASES

NUMBER OF CASES

60

The above analysis shows the performance of Broselow tape in estimating

weight of the children, in which only 47.7% of children have predicted weight

within 10% of actual weight. 69.6% cases are within 15% of actual weight. From

our analysis, the prediction accuracy of the Broselow tape in estimating weight

of children was not significant in any of the groups (p<0.0001). In this study

population, Broselow tape did not perform well in estimating weight.

61

DISCUSSION

Safety of patients and reducing the medical errors, both, are contemporary

issues in emergency settings(58). In our settings, we have found that ED has been

the most error prone place with drug dosage error as the principal cause and

paediatric age group as the most common patients(59).

In an emergency resuscitation situation, management of critically ill

children requires accurate weight to determine drug doses. Hence we need a

reliable method of estimating weight in that particular emergent setting. The end

point in estimation of weight, therefore, should be more than just the weight itself,

which means during management of sick children, the resuscitation team must

concentrate on medical management of patient and should not get distracted by

other issues such as determination of drug doses and equipment sizes which can

be made easy by incorporating evidence based resuscitation aids such as looking

them on a chart or computer that does not require active or non-automatic

thinking or memory. In view of this simple reason, a complete system is ideal for

use during the paediatric resuscitation so that one or more devices can be used

automatically to predict weight, to guide on equipment size selection and to

provide drug dosaging, dilution and delivery information. Other issues that are

important in the selection or development of a resuscitation aid include cost,

availability, ease of use, disposability and transportability.

There are 2 components in a resuscitation aid. First one being the weight

estimation system, which are best embodied in a length-based tape system. Even

62

these tape based systems are vulnerable to errors because of incorrect

measurement or inaccurate weight estimation. The second component is to

provide essential information to the resuscitation team on drug doses and

equipment sizes which should not be trusted to memory. It may come from a tape

or a chart or a booklet that should be specifically designed to provide relevant

information to the entire team and not just the doctor to minimise the possibility

of errors during a time of high cognitive load and emotional stress. This aspect of

the resuscitation aid has not been addressed in any way in this study.

Currently Broselow tape is used during resuscitation aid and this tape have

incorporated both the components. The discussion of the concept of resuscitation

aids is important because the design of a weight estimation system should be

determined largely by its ability to improve patient care and not just provide an

estimate of weight. Our study is a validation study and has just taken weight

estimation part alone and has not touched regarding accuracy of drug dosaging

and equipment sizes.

Previous studies have used target estimation error of within 10% as

acceptable for drug dose calculation and this error percentage has been followed

by most in weight estimation studies(60). Few studies have been done using a

weight-estimation error of within 20% as acceptable(61). Our study also have used

10% error as acceptable for drug dose calculation. The reason why an error

percentage of 10 has been allowed was other latent, unavoidable and undetectable

errors may occur. It is outrageous to expect weight estimation to be more precise

63

than the drug dose error that is already possible because of drug concentration

variability. Errors related to the rounding-off of body weight also contributes to

a remarkable degree of variability that cannot easily be determined.

In our study, Broselow tape estimated weight of 47.7% of the study

population within 10% of a child’s measured weight. This statistics was found to

be lower when compared to other studies. The outcome shows that more than half

of the study population had a weight estimate with an error exceeding 10%. In a

study by Saily Britnell et al, they have found that Broselow tape included 73.4%

of children within 10% of a child’s measured weight. In a study by Michael David

John Wells et al, the percentage of children within 10% of measured weight was

around 64%. From this comparison, it is evident that Broselow tape did not

perform well in our study population. The number of children in this study with

a weight estimation of ≤10% error was lower and the number of our children with

a weight estimation error of >15% was also greater. Similar negative findings

were shown in studies done by Ramarajan N et al, Kun W et al, Theron l et al.

In our study, Broselow tape underestimated weight by 1.6 kg, where as in

a study by Michael David John Wells et al, Broselow tape underestimated weight

by 0.9 kg. In a study by Farhad Asskaryar et al, they found that Broselow tape

overestimates weights with a 5–15 % mean percentage difference in the Indian

children and misclassifies them into a wrong color zone on the Broselow tape,

usually to a higher color zone.

64

The variations may be due to varied number of reasons. In Indian children,

there are several possible factors for this trend. Level of nourishment, exclusive

breast feeding, low birth weight, socio-economic factors are the possible

contributory factors to highlight current data and their relevance to lag in ideal

height for weight and weight for age in the indian population.

In a recent survey in southern India, where our study was conducted, one

study has determined that the prevalence of undernutrition (≤80 percentage of

standard weight for age) was 66.5 %. Such prevalence tends to increase with age

and there is a stark difference between male (76.9 %) and female (56.3 %)

genders. The prevalence of grade 1 malnourishment was 46.2 %. The prevalence

decreased as the socioeconomic status improved and the duration of EBF had an

influence on the nutritional status of the growing child(62).

The coefficient correlation of the agreement between measured weight and

weight predicted by Broselow tape for entire study population and for each

groups are comparable with other studies. Our study had a correlation coefficient

of 0.97. The study by Michael David John Wells et al had a correlation coefficient

of 0.946 and in the study by Farhad Asskaryar et al the correlation coefficient was

found to be 0.954 for boys and 0.9 for girls.

65

TABLE 14: COMPARISON OF CORRELATION COEFFICIENT

STUDY

CORRELATION COEFFICIENT

OUR STUDY

0.97

Michael David John Wells et

0.946

Farhad Asskaryar et al

0.954 for boys

0.9 for girls

Varghese A et al

0.974

The other major indicator of bias, the mean percentage error (MPE) of our

study was 8.5% when compared to the study done by Michael David John Wells

et where the MPE was 3.8%.

In our study, there was no significant difference in the performance of the

tape when males and females were compared which was again similar to other

studies. Here the precision was more or less identical between genders but the

bias towards underestimation in the Broselow tape was less in female children

when compared to male children.

66

The reasons why the tape has not performed better in our population may

possibly be the use of NCHS 2000 weight for length charts by the Broselow tape

and the positioning of the child in the tape because the divisions in Broselow tape

occupy the area below the 50th centile which could have made the tape to

underestimate weight in children.

67

CONCLUSION

The correlation with actual weight was better overall and in each weight

category.

Both measures of the tape bias (Mean Percentage Error and Bland-Altman

analysis) were significantly higher in the study population.

The measure of tape precision (SD of Mean Percentage Error) was

significantly worse in the study population.

Broselow tape predicted weight within 10% of actual weight only in 47.7%

of the entire study population.

The results was similar in each weight category. In 10-20 kg weight

category, tape predicted weight within 10% ofn actual weight in over 60%

of children but it was not statistically significant to prove that the tape was

useful in estimating weight in that category.

Broselow tape often underestimates weight in our study population.

Statistical analysis shows that the tape has underestimated weight by 1.6

kg.

The Mean percentage error was 8.5%, which is high. SD of MPE was

11.49.

Broselow tape did not perform well in our study population.

The implication of this study is that Broselow tape is relatively inaccurate

and that with increasing age and weight they become more inaccurate.

68

Further work is required to revise the methods by which weight estimation

is performed;

Additional research is urgently needed to establish a pharmacokinetic

“target weight” which should be used for drug dose calculation, and

therefore weight estimation strategies

69

STRENGTH OF THIS STUDY

This is a prospective study to validate an existing instrument (tape).

The study population was exclusively used for validation of the instrument

(tape) and did not contribute to the development of new instrument.

This study was adequately powered to allow for accurate equivalence

within 4 subgroups.

The actual tape was used to obtain measurement rather than a tape measure.

Children were measured in supine position to obtain a true length rather

than height.

Bland-Altman plot and Mean Percentage Error were used in analysis as a

measurement of bias and precision.

70

LIMITATIONS

Our study is a single centre prospective study.

Although our study sample is diverse and representative of all geographic

regions and socioeconomic backgrounds in the Indian population, it is

collected at one tertiary clinic.

There is no comparable NCHS data pool available in India that would allow a

precise estimate of height and weight for each color zone.

In this study, we have considered only gender difference, and other categories

such as social class, income, parent education level, differences in ethnicity, and

regional areas are not considered in the analysis. We could not with certainty verify the age of the children in our data

sample and relied on the information obtained from the parents or guardian. Children requiring emergency treatment were excluded and this may have

introduced some bias in two respects: children with critical illness or injury

may be different to ambulant paediatric patients in terms of their body

composition; and the usual physical, psychological and cognitive stresses

experienced by healthcare personnel during a paediatric resuscitation were

not present. Our study is a validation study and did not include other features of the

tape such as testing the dosages and equipment sizes whether or not

appropriate for the study population.

71

RECOMMENDATIONS

The Broselow tape needs to be evaluated in a number of different

populations to establish its accuracy across a range of socio-economic

strata and ethnic and racial characteristics.

72

HEIGHT OF INDIAN CHILDREN (BOYS AND GIRLS AGED 0-18 YRS)

73

WEIGHT OF INDIAN CHILDREN (BOYS AND GIRLS AGED 0-18

YEARS)

74

BIBLIOGRAPHY

1. Kaushal R, Bates DW, Landrigan C, 2001. Medication errors and adverse

drug events in pediatric inpatients. JAMA, 285:2114-2120.

2. Shah AN, Frush K, Luo X, Wears RL. Effect of an intervention

standardization system on pediatric dosing and equipment size

determination: a crossover trial involving simulated events. Arch Pediatr

Adolesc Med 2003; 157: 934.

3. Argall JAW, Wright N, Mackway-Jones K, Jackson R, 2003. A

comparison of two commonly used methods of weight estimation. Arch

Dis Child, 88:789–790.

4. Krieser D, Nguyen K, Kerr D, Damien J, Clooney M, Anne-Maree K,

2007. Parental weight estimation of their child's weight is more accurate

than other weight estimation methods for determining children's weight in

an emergency department? Emerg Med J, 24(11):756-759.

5. Varghese A, Vasudevan VK, Lewin S, Indumathi CK, Dinakar C, Rao SD,

2006. Do the Length-based (Broselow) Tape, APLS, Argall and Nelson’s

formulae accurately estimate weight of Indian children? Indian Pediatr,

43(17):889-894.

6. Black K, Barnett P, Wolfe R, Young S, 2002. Are methods used to estimate

weight in children accurate? Emerg Med, 14:160–165.

75

7. Lubitz DS, Seidel JS, Chameides L, Luten RC, Zaritsky AL, Campbell FW,

1988. A rapid method for estimating weight and resuscitation drug doses

from weight in the pediatric age group. Ann Emerg Med, 17(6):576-81.

8. Mackway-Jones K, Molyneux E, Phillips B, Wieteska S, 2001. Advanced

pediatric life support: The practical approach. 3rd Ed. London: BMJ

Publishing Group, p: 7.

9. Vilke GM, Marino A, Fisher R, Chan TC, 2001. Estimation of paediatric

weight by EMT-PS. J Emerg Med, 21:1125–1128.

10. Tinning KK, Acworth JP, 2006. Make your best guess: an updated method

for paediatric weight estimation. Proceedings of Annual Scientific

Meeting. Australasian College for Emergency Medicine.

11. DuBois D, Baldwin S, King WD, 2007. Accuracy of weight estimation

methods for children. Pediatr Emerg Care, 23(4):227-230.

12. Luscombe M, Owens B, 2007. Weight estimation in resuscitation: is the

current formula still valid? Arch Dis Child, 92(5): 412–415.

13. Theron L, Adams A, Jansen K, Robinson E, 2005. Emergency weight

estimation in Pacific Island and Maori children who are large-for-age.

Emerg Med Australas, 17:238-243.

14. Nguyen K, Krieser D, Kerr D, Jolley D, Clooney M, Kelly AM, 2007.

Failed validation of the Argall weight formula for estimating children's

weight in an Australian ED. Acad Emerg Med, 14(5):486-488.

76

15. Hofer CK, Ganter M, Tucci M, Klaghofer R, Zollinger A. How reliable is

length-based determination of body weight and tracheal tube size in

pediatric age group? The Broselow tape reconsidered. Br J Anaesth 2002;

88: 283- 285.

16. Luten RC, Wears RL, Broselow J, Zaritsky A, Barnett TM, Lere T, et al.

Length-based endo-tracheal tube and emergency equipment in pediatrics.

Ann Emerg Med 1992; 21:900-904.

17. Bates DW, Cullen D, Laird N, et al. Incidence of adverse drug events and

potential adverse drug events: implications for prevention. JAMA.

1995;274:29-34.

18. Nelson textbook of pediatrics; 20th edition, kleigman, Stanton, St Geme,

Schor.

19. Broselow J, inventor, Luten R, inventor Vital Signs, Inc, assignee.

Broselow Pediatric Emergency Tape. US patents 4 713 888, December 22,

1987; 4 823 469, April 25, 1989; 5 010 656, April 30, 1991.

20. Ferner, R. E. (2012). Medication errors. British Journal of Clinical

Pharmacology, 73(6), 912-916. doi:10.1111/j.1365-2125.2012.04236.

21. Park, J., Kwak, Y. H., Kim do, K., Jung, J. Y., Lee, J. H., Jang, H. Y., . . .

Hong, K. J. (2012). A new age-based formula for estimating weight of

Korean children. Resuscitation, 83(9), 1129-1134.

doi:10.1016/j.resuscitation.2012.01.023.

77

22. Cattermole, G. N., Leung, M. P., So, H. K., Mak, P. S., Graham, C. A., &

Rainer, T. H. (2011). Age-based formulae to estimate children's weight in

the emergency department. Emergency Medicine Journal, 28(5), 390-396.

23. Lulic, I., & Kovic, I. (2010). Comparing methods for weight estimation of

children: A preliminary study. Resuscitation, 81(2), S77.

doi:10.1016/j.resuscitation.2010.09.315.

24. Sandell, J. M., & Charman, S. C. (2009). Can age-based estimates of

weight be safely used when resuscitating children? Emergency Medicine

Journal, 26(1), 43-47. doi:10.1136/emj.2008.061119.

25. Geduld, H., Hodkinson, P. W., & Wallis, L. A. (2011). Validation of

weight estimation by age and length based methods in the Western Cape,

South Africa population. Emergency Medicine Journal, 28(10), 856-860.

doi:10.1136/emj.2010.098640.

26. Abdel-Rahman, S. M., Ahlers, N., Holmes, A., Wright, K., Harris, A.,

Weigel, J., . . . Kearns, G. L. (2013a). Validation of an improved pediatric

weight estimation strategy. The Journal of Pediatric Pharmacology and

Therapeutics, 18(2), 112-121. doi:10.5863/1551-6776-18.2.112

27. Advanced Life Support Group. Advanced paediatric life support: the

practical approach. 4th edn. London: BMJ Publishing Group, 2005.

28. Hamill PV, Drizd TA, Johnson CL, et al. NCHS growth curves for children

birth-18 years: United States. Vital Health Stat 11 1977;165: 174.

78

29. Carrol W, Alexander J. Towards a better weight estimation in the seriously

ill child - a comparison of methods. Arch Dis Child 2001;84(Suppl I):A12.

30. Weech A. Signposts on the highway of growth. AJDC 1956;88:452-457.

31. Traub SL, CE J. Comparison of methods of estimating creatinine clearance

in children. Am J Hosp Pharm 1980;37:195-201.

32. Traub SL, Kitchen L. Estimating ideal body mass in children. Am J Hosp

Pharm 1983;40:107-110.

33. Garland JS, Kishaba RG, Nelson DB, Losek JD, Sobocinski KA. A rapid

and accurate method of estimating body weight. Am J Emerg Med

1986;4:390-393.

34. Oakley P. Inaccuracy and delay in decision making in paediatric

resuscitation, and a proposed reference chart to reduce error. BMJ

1988;297:817-819.

35. Leffler S, Hayes M. Analysis of parental estimates of children's weights in

the ED. Ann Emerg Med 1997;30:167-170.

36. Hazell W. Emergency medicine at your fingertips. 2nd ed. New Zealand:

Medimedia (NZ); 2000; 132

37. Lo C, Chan K, Chik S. A tape measure to aid paediatric resuscitation -

modification for local use. HKJEM 1995;2(1):51-57.

38. Molyneux E, Brogan R, Mitchell G, Gove S. Children's weights: guess or

measure by tape. Lancet 1999;354:1616.

39. Kloeck W. The Kloeck tape. Personal communication 2008.

79

40. Sandell J. The Sandell tape. 2004 [cited 2008 11 November]; Available

from: www.sandelltape.com.

41. Brierly J. The paediatric resuscitation emergency management system.

2005 [cited 2008 11 November]; Available from: www.premsystem.co.uk.

42. Nieman CT, Manacci CF, Super DM, Mancuso C, Fallon WF. Use of the

Broselow tape may result in the underresuscitation of children. Acad

Emerg Med 2006;13:1011–1019.

43. Ramarajan N, Krishnamoorthi R, Strehlow M, Quinn J, Mahadevan SV.

Internationalizing the Broselow tape: how reliable is weight estimation in

Indian children. Acad Emerg Med 2008;15:431–436.

44. Kun W, Cheng K, Yuen M, Tung W. How good is the Broselow tape

measurement for estimation of body weights in paediatric patients for

application in Hong Kong? HK J Paediatr 2000;5(1):25-30.

45. Dieckmann RA. The dilemma of paediatric drug dosing and equipment

sizing in the era of patient safety. Emerg Med Australas 2007;19:490-493.

46. Deboer S, Seaver M, Broselow J. Color coding to reduce errors. Am J Nurs

2005;105(8):68-71.

47. Thompson MT, Reading MJ, Acworth JP. Best Guess method for agebased

weight estimation in paediatric emergencies validation and comparison

with current methods. Emerg Med Australas 2007;19:535-542.

80

48. Kelly A-M, Kerr D, Clooney M, Krieser D, Nguyen K. External validation

of the Best Guess formulae for paediatric weight estimation. Emerg Med

Australas 2007;19:543-546.

49. Banach A, Wade TJ, Cairney J, Hay JA, Faught BE, O'Leary DD.

Comparison of anthropometry and parent-reported height and weight

among nine year olds. Can J Public Health 2007;98(4):251-253.

50. Goldman RD, Buskin S, Augarten A. Parental estimates of their child's

weight: accurate for resuscitation drug doses. Pediatr Emerg Care

1999;15(1):19-21.

51. Dearlove CB, Dearlove O. Visual estimation of children's weights.

Anaesthesia 1999;54(12):1228-1229.

52. DuBois D, Baldwin S, King WD. Accuracy of weight estimation methods

for children. Pediatr Emerg Care 2007;23(4):227-230.

53. Menon S, Kelly AM. How accurate is weight estimation in the emergency

department? Emerg Med Australas 2005;17(2):113-116.

54. Nash A, Corey M, Sherwood K, Secker D, Joanne Saab, OConnor DL

Growth assessment in infants and toddlers using three different reference

charts. J Pediatr Gastroenterol Nutr 2005;40:283-288.

55. Nash A, Secker D, Corey M, Dunn M, O’Connor DL. Field testing of the

2006 World Health Organisation growth charts from birth to 2 years -

assessment of hospital undernutrition and overnutrition rates and the

usefulness of BMI. J Parent Ent Nutr 2008;32(2):145-153.

81

56. Haftel AJ, Khan N, Lev R, Schonfeld N. Hanging leg weight -- a rapid

technique for estimating total body weight in pediatric resuscitation. Ann

Emerg Med 1990;19(5):523-526.

57. Bavdekar SB, Sathe S, Jani P. Prediction of weight of Indian children aged

up to two years based on foot-length: implications for emergency areas.

Indian Pediatr 2006;43(2):125-130.

58. Kohn L, Corrigan J, Donaldson M. To err is human: building a safer health

system. IOM, editor. Washington, DC: National Academy Press; 2001;

59. Cohen AL, Budnitz DS, Weidenbach KN, et al. National surveillance of

emergency department visits for outpatient adverse drug events in children

and adolescents. J Pediatr 2008;152:416-421.

60. Nieman CT, Manacci CF, Super DM, Mancuso C, Fallon WF. Use of the

Broselow tape may result in the underresuscitation of children. Acad

Emerg Med 2006;13:1011–1019.

61. Thompson MT, Reading MJ, Acworth JP. Best Guess method for agebased

weight estimation in paediatric emergencies validation and comparison

with current methods. Emerg Med Australas 2007;19:535-542.

62. NFHS-3 2005–2006. http://www.rchiips.org/nfhs. Accessed April 05,

2015.

PROFORMA

NAME :

AGE :

SEX :

ADDRESS :

OP/IP NO :

ANTHROPOMETRY

WEIGHT (ACTUAL) :

WEIGHT (BROSELOW) :

HEIGHT :

MAC :

ABBREVIATIONS

APLS - Advanced Paediatric Life Support

ED - Emergency Department

MPE - Mean percentage Error

NCHS - National Center for Health Statistics

OPD - Out Patient Department

PAWPER - Paediatric Advanced Weight Prediction in the

Emergency Room

PREM - Paediatric Resuscitation Emergency Management

SD - Standard Deviation

SPSS - Statistical Package for the Social Sciences

Name Age M-1 F-2 Actual Wt Group Tape Wt Group krishnan 4 1 12 2 12 2arun 5 1 15 2 14 2shasthika 1 2 7 1 8 1kiruba 3 1 11 2 10 1kirthi 7 2 17 2 16 2kannan 10 1 28 3 30 3niranjan 5.5 1 16 2 17 2mani 0.5 1 5 1 6 1manikanda 9 1 25 3 24 3jakan 12 1 30 3 32 4janani 10 2 32 4 36 4sowmiya 8 2 29 3 32 4sanjana 2 2 10 1 12 2saran 2.5 1 11 2 12 2aaradhna 2 2 9 1 10 1kamal 1 1 7.5 1 8 1ajish 1.5 1 8.5 1 8 1lalitha 7 2 20 2 20 2kiran 6 1 23 3 22 3suresh 5 1 20 2 20 2vinoth 5 1 21 3 22 3karthikeyan 5 1 27 3 30 3mahesh 9 1 30 3 34 4madhan 8 1 28 3 32 4viji 10 2 35 4 38 4kesavan 11 1 39 4 38 4murali 6 1 23 3 22 3pushkar 4.5 1 20 2 22 3pooja 8 2 30 3 36 4kirthi 8.8 2 28 3 34 4keerthana 9 2 31 4 36 4nandhini 9 2 33 4 38 4ilakia 11 2 32 4 38 4kavitha 2 2 11 2 12 2vanitha 1.5 2 9 1 10 1rajkamal 3.5 1 14 2 14 2rajesh 6 1 20 2 20 2arul 8 1 30 3 26 3paandi 7 1 28 3 30 3soundar 4 1 15 2 14 2vijay 3.5 1 13 2 12 2anurega 4 2 14 2 14 2aravind 2.5 1 11 2 12 2sumathi 6 2 18 2 16 2karuppu 8 1 25 3 24 3muthu 4 1 13 2 12 2b/o rani 0.1 2 4 1 4 1rahul 5 1 17 2 16 2sudharsan 4.7 1 15 2 14 2

gokul 6.2 1 20 2 18 2vasanth 6.7 1 21 3 20 2thomas 7 1 24 3 26 3nilofer 8 2 27 3 28 3asmath 7.5 1 25 3 26 3asmaas 7 1 26 3 30 3amul 3 1 13 2 12 2raja 7 1 20 2 20 2rajashri 9 1 27 3 30 3vasanthi 10 2 31 4 36 4kalpana 11 2 33 4 38 4suganya 6 2 20 2 18 2anitha 3 2 11 2 12 2kannan 2 1 10 1 10 1karthikeyan 5 1 15 2 14 2rani 2 2 11 2 12 2durai 2.5 1 14 2 12 2gowtham 8 1 23 3 26 3shoban 6 1 20 2 24 3harish 2 1 12 2 10 1nithish 1.5 1 9 1 10 1nirupama 1 2 8 1 8 1claret 1 1 8.5 1 8 1anton 3 1 13 2 12 2suganthi 3 2 13.5 2 12 2suthan 3.5 1 15 2 14 2lohith 7 1 20 2 18 2sujith 5 1 18 2 18 2aathi 3 1 14 2 14 2nagul 2 1 11 2 12 2vijay 5 1 18 2 18 2ajith 9 1 31 4 36 4setupathi 2.5 1 12 2 12 2kumaran 1.5 1 9 1 10 1jana 0.5 1 4.5 1 6 1balaji 2 1 11 2 10 1arunan 4 1 17 2 16 2valli 5 2 16 2 16 2petchi 6 2 20 2 18 2karuppan 7 1 21 3 24 3kanmani 10 2 35 4 40 4kishore 2 1 11 2 12 2vignesh 5 1 19 2 18 2varathan 1 1 9 1 10 1senthil 0.3 1 6 1 6 1kanthan 3 1 13 2 12 2vishal 0.5 1 5.5 1 6 1surya 7 1 20 2 18 2shtuthi 4 2 15 2 14 2venmathi 8 2 22 3 20 2

girija 2 2 12 2 12 2suganya 4 2 16 2 16 2thangam 9 2 30 3 38 4veena 11 2 29 3 36 4vanthana 6 2 21 3 20 2selvam 4.5 1 18 2 16 2selvi 6.5 2 25 3 24 3rajgopal 3 1 12 2 12 2ramanan 1 1 8 1 10 1rajiv 3 1 13 2 12 2anbu 2.5 1 14 2 14 2rajaram 6 1 24 3 22 3joseph 7 1 23 3 24 3john 5 1 20 2 24 3christy 1 1 8 1 10 1jameela 1.5 2 9 1 8 1mohd.sam 3 1 12 2 10 1kiruba 3.5 2 12 2 10 1sameera 6 2 23 3 22 3diana 9 2 28 3 34 4daisy 2 2 11 2 10 1ganesh 3.5 1 15 2 14 2perumal 1 1 8 1 8 1ramu 1 1 7.7 1 10 2rasiga 1.5 2 9.4 1 8 1santhi 3 2 13.6 2 16 2jeyanthi 4 2 15 2 14 2jaya 6 2 18 2 22 3karnan 5.5 1 17 2 18 2gopal 5 1 17 2 16 2harish 2 1 11 2 10 1jagadesh 2.5 1 13 2 12 2govarthan 6 1 18 2 16 2girija 5.5 2 17 2 14 2girish 4 1 16 2 16 2saran 1 1 9 1 10 2sarath 1.3 1 10 1 10 2saratha 8 2 28 3 34 4saravanan 10 1 30 3 34 4tamizh 4 1 17 2 16 2nagulan 3 1 15 2 14 2surendar 4 1 16 2 14 2sarathy 5.5 1 18 2 20 2priya 8 2 26 3 34 4priya 2 2 12 2 12 2nisha 1 2 7.9 1 8 1jeslin 3 2 14.7 2 12 2bakrudhen 12 1 36 4 40 4jayakumar 2 1 13 2 12 2rajavel 6 1 19 2 20 2

kumar 3 1 14 2 14 2vennila 2 2 11 2 12 2varun 1.5 1 10 1 10 2bharathi 7 2 22 3 22 3ilamaran 4 1 18 2 16 2seeman 9 1 32 4 36 4ramasamy 2 1 12 2 12 2rajathi 2.5 2 13 2 12 2rajan 6 1 25 3 22 3vaishnavi 4 2 19 2 26 3vasikaran 3.5 1 15 2 14 2vaishali 1 2 9 1 8 1vaijayanthi 0.5 2 6 1 8 1karthiga 10 2 Feb-00 4 40 4kirthika 7 2 27 3 32 4prasanna 8 1 26 3 30 3prasath 1 1 8.5 1 8 1prabakar 1.5 1 9.5 1 10 1pranesh 1 1 8 1 10 1prema 1 2 9 1 10 1pradeep 0.5 1 5.5 1 6 1sankar 11 1 33 4 36 4ramya 12 2 38 4 40 4ramesh 2 1 11 2 12 2rajini 3 1 14 2 14 2siva 5 1 20 2 20 2nagarajan 6 1 24 3 22 3sundar 8 1 29 3 36 4sudhir 2 1 11 2 12 2arunan 1 1 9 1 10 1aruna 1.5 2 11 2 12 2rajeev 3.5 1 16 2 16 2pachamuth 4.5 1 18 2 20 2sivnesh 2.5 1 14 2 14 2sivakaran 3 1 17 2 18 2muzzamil 2 1 12 2 12 2sanjay 1 1 9 1 10 1kicha 1.5 1 11 2 12 2sham 3 1 15 2 16 2harini 6 2 21 3 24 3paapu 8 2 30 3 36 4padma 9 2 32 4 38 4lacanya 3 2 12 2 12 2latchana 5 2 17 2 20 2lagubaran 7 1 21 3 26 3mano 10 1 33 4 38 4manoj 2 1 12 2 12 2manoharan 2 1 12 2 12 2pallavan 1 1 9 1 10 1harshini 3 2 14 2 16 2

jothi 2 2 12 2 12 2vetrivel 2.5 1 12 2 14 2yashvanth 1.5 1 10 1 10 1nithish 6 1 22 3 24 3thulasi 1.5 2 10 1 12 2anushri 1 2 8 1 10 1rithesh 1.5 1 11 2 12 2arun 1.5 1 11 2 12 2peadeep 2.5 1 14 2 12 2devandran 2 1 10 1 12 2pandi 4 1 15 2 16 2shanmugam 7 1 20 2 22 3naveen 2.5 1 14 2 16 2abdul 5 1 22 3 22 3arivarasan 8 1 30 4 34 4siva 12 1 36 4 40 4dharsini 0.5 2 5.5 1 6 1tarun 4 1 16 2 16 2bathrakaal 4.5 2 17 2 16 2ashwin 3 1 11 2 12 2vennila 2.5 2 12 2 12 2kanmani 8 2 29 3 34 4yaashrai 6 1 25 3 28 3sanjana 4.5 2 20 2 22 3rishwana 8 2 28 3 34 4martin 9 1 30 3 34 4vimal 1 1 9 1 10 1sasi 2 1 12 2 12 2dinesh 12 1 35 4 40 4swetha 2.5 2 15 2 16 2sham 5 1 19 2 17 2lingam 11 1 32 4 36 4vicky 5 1 20 2 22 3rajpriya 7.5 2 28 3 34 4ashwin 6 1 28 3 36 4ramani 0.5 2 5 1 6 1john 9 1 29 3 34 4suber 12 1 38 4 42 4b/o ramya 0.2 2 4 1 4 1mumtaz 11 2 31 4 34 4sanish 2 1 12 2 12 2kirmani 1 1 9 1 10 1dhoni 11 1 32 4 42 4dharman 3 1 14 2 12 2murugesh 4 1 17 2 14 2mayil 8 1 29 3 36 4kumar 5 1 20 2 22 3revathy 0.7 2 7.5 1 10 1revathi 2 2 12.5 2 12 2ramprabhu 1.5 1 11 2 12 2

vicky 1.5 1 10.5 2 12 2logesh 2.5 1 13 2 12 2logeshwara 3 1 12.5 2 12 2pranesh 4 1 15 2 14 2suresh 8 1 27.5 3 32 4valavan 9 1 30 3 32 4raghuvaran 1 1 8.5 1 10 1guru 0.4 1 5 1 6 1rahavan 4 1 16 2 16 2prabhu 3 1 14 2 14 2karthikeyan 7 1 24 3 28 3vikram 9 1 29 3 34 4dhruva 8 1 28 3 32 4santhosh 5 1 20 2 22 3chandru 4 1 17 2 20 2chandran 7 1 26 3 30 3kumari 3 2 14 2 15 2kaanchana 6.5 2 24 3 28 3cinthiya 2 2 11 2 12 2sindhu 1.5 2 10.5 2 12 2sindhuja 12 2 36 4 42 4vanthana 11 2 35 4 42 4vatchala 10 2 35 4 42 4varuni 9 2 32 4 42 4dhivya 8 2 27 3 34 4manohari 7 2 28 3 34 4sudharsan 6 1 25 3 30 3velmuruga 5 1 20 2 22 3palani 4 1 17.6 2 20 2vijay 4.5 1 22 3 24 3sethu 3 1 16 2 20 2karupan 3.5 1 15 2 18 2goutham 2.5 1 12 2 12 2sabari 2 1 10.5 2 12 2visu 1.5 1 11 2 12 2jakan 1 1 8.5 1 10 1kesavan 0.5 1 5.5 1 6 1pethukani 0.5 2 5.5 1 6 1dorian 2 1 11 2 12 2selvakuma 3.5 1 13 2 14 2selvaraj 5 1 18 2 20 2kamesh 3 1 13 2 15 2rajkumar 2.5 1 12 2 14 2balaji 3.5 1 12 2 14 2bala 3 1 14 2 12 2balan 1 1 8 1 10 1ganapathy 6 1 21 3 20 2vaani 7 2 25 3 20 2laxmi 3 2 13 2 12 2laxmanan 8 1 30 3 26 3

sasivarnan 9 1 27 3 24 3vetrivel 2 1 12 2 10 1yashvanth 1 1 8 1 7 1nithish 1.5 1 10 1 12 2thulasi 3 2 14 2 14 2kanika 6 2 19 2 20 2kanishka 4 2 16 2 18 2anisha 8 2 29 3 32 3anusha 9 2 28 3 36 4arun 10 1 35 4 42 4pradeep 12 1 38 4 44 4mano 2.5 1 13 2 14 2manoj 4.5 1 20 2 20 2manish 6.5 1 23 3 24 3devenran 8 1 28 3 32 4pandi 7 1 25 3 28 3parthi 2 1 12 2 14 2shanmugam 1.5 1 9 1 10 1martin 0.5 1 5 1 6 1vimal 12 1 31 4 44 4rahman 11 1 28 3 42 4arivarasan 3 1 13 2 14 2siba 2.5 1 12 2 12 2dharsini 5 2 16 2 16 2bathrakali 6 2 19 2 20 2ashwin 4 1 18 2 20 2vennila 8 2 30 3 36 4kannika 9 2 27 3 30 3kaamesh 12 1 33 4 38 4rasool 4.5 1 19 2 20 2moideen 4.5 1 17 2 20 2arul 6 1 20 2 20 2sasi 7 1 21 3 22 3dinesh 8 1 25 3 24 3swetha 9 2 29 3 34 4sham 2 1 10 1 12 2lingam 1 1 8.5 1 10 1lingesh 1.5 1 10 1 10 1rajpriya 1.5 2 12 2 12 2rampaul 1 1 7.9 1 10 1mumtaz 0.5 2 4.5 1 6 1abirami 1 2 8 1 10 1anton 1 1 8.3 1 10 1vaishali 2 2 12.5 2 14 2jana 2.5 1 13.5 2 16 2balaji 3.5 1 16 2 16 2prasanna 4.5 1 15.6 2 16 2viji 5.5 2 20 2 24 3b/o guna 6 2 22.5 3 24 3ramani 7 2 21 3 24 3

aneesh 6.5 1 22 3 24 3sujith 3 1 14 2 16 3sharav 5 1 17 2 20 2abhinav 8 1 27 3 32 3abhinesh 9 1 28 3 34 4elangovan 9.5 1 30 3 34 4dalton 5 1 18 2 20 2kingston 6 1 20 2 24 3satya 2.5 1 13 2 14 2sankar 3 1 14 2 16 2srinivasan 1.5 1 11 2 12 2sivakaran 10 1 33 4 40 4prathap 9 1 32 4 40 4pranesh 4 1 14 2 16 2jagadesh 5.5 1 19 2 22 3govarthan 3 1 13 2 14 2navaneeth 4 1 15 2 16 2udhay 3.5 1 13 2 14 2ilankumara 2 1 11 2 12 2jayaprakas 2.5 1 12 2 14 2rajkamal 2 1 12 2 14 2michael 6 1 19 2 22 3daniel 7 1 23 3 24 3harish 5.6 1 27 3 30 3najih 6 1 24 3 26 3sowmiya 2 2 13 2 14 2aravind 1.5 1 10 1 12 2ravichandr 1 1 6.9 1 10 1menaka 0.5 2 4.8 1 6 1kalai 0.6 2 5.2 1 6 1kovalan 0.3 1 4.4 1 6 1senthil 1 1 8.7 1 10 1velavan 1 1 9.2 1 10 1mani 0.2 1 4 1 6 1valluvan 2 1 13.5 2 14 2kamaraj 3 1 13 2 14 2devar 1 1 9 1 10 1palani 4 1 16 2 18 2logesh 3.5 1 14.5 2 16 2sivakumar 8 1 32 4 38 4kannaki 7 2 30 3 38 4salman 6 1 28 3 34 4eswaran 11 1 32 4 40 4kamalesh 2 1 12 2 14 2tamizh 2.5 1 14 2 14 2pugazh 3 1 13 2 14 2aadhi 3.5 1 15 2 16 2arul 7 1 22 3 24 3murugan 7.5 1 24 3 26 3ramanan 8 1 28 3 32 4

ramana 2 1 11 2 14 2kishore 9 1 24 3 26 3veeraraj 10 1 23 3 24 3nagarathna 0.2 2 4 1 4 1kaaleswari 0.7 2 8 1 10 1atchaya 1.5 2 10 1 10 1sanmitha 0.5 2 5.5 1 6 1kathirvel 0.6 1 7 1 8 1gubendran 3 1 11 2 10 1syed 10 1 20 2 32 3maheswari 5 2 15 2 14 2aswanth 7.5 1 22 3 24 3varmith 0.4 1 4 1 4 1vignesh 12 1 29 3 32 4dharsan 0.8 2 6 1 6 1santhosh 5.5 1 13 2 14 2madhan 11 1 36.4 4 42 4vasanth 7 1 15.7 2 18 2ratchan 3 1 11 2 12 2thangesh 0.6 1 7.6 1 8 1mithun 12 1 23 3 24 3harishkuma 7 1 18 2 20 2sabari 0.3 1 4.5 1 6 1veeranan 6 1 19.2 2 22 3bhuvana 3 2 11 2 12 2sivasankar 0.8 1 18.2 2 20 2kamal 0.7 1 7.6 1 8 1vaishnavi 6 2 16.8 2 18 2arulkumara 6 1 18 2 20 2kishorepan 7 1 15 2 16 2sabaripriya 6 2 13.7 2 14 2udhaykuma 7 1 18.6 2 20 2tina 3 2 12.5 2 14 2diwakar 6 1 17.4 2 18 2yogeshwar 11 1 28.4 2 32 4mareeswar 8 2 20 2 28 3hansika 6 2 17.5 2 22 3ezhilarasan 4.5 1 15.1 2 18 2azrina 10 2 28.7 3 34 4nivas 8 1 28 3 28 3shenbagav 12 2 35.7 4 44 4sonasri 1.5 2 8.3 1 10 1poorvika 3 2 12.5 2 16 2sandyari 8 2 16.9 2 18 2sibana 1.5 2 10 1 12 2rithik 2.5 1 12.5 2 14 2dasarathan 3.5 1 12.5 2 14 2nagalaxmi 12 2 23.9 3 28 3nithish 8 1 22 3 26 3bhuvana 3.5 2 11.4 2 14 2

vimalesh 2 1 10.8 2 12 2riyaskhan 7 1 15.9 2 18 2sujai 5 1 14.8 2 18 2lisi 7 2 22.1 3 24 3dharika 5 2 15 2 18 2harishwara 7 1 18.2 2 20 2samitha raj 4.5 2 13 2 14 2gokulkanna 3.5 1 8.8 1 10 2muganthes 2 1 10.5 2 12 2krithika 9 2 25 3 24 3senthilkum 2.5 1 8.9 1 10 1pattaveera 11 1 22.7 3 24 3saravanaku 8.5 1 17.4 2 18 2muthukavit 6.5 2 13.6 2 14 2akashpand 4.5 1 14 2 14 2dinesh 7 1 12.4 2 14 2yuvaraj 1.5 1 10.1 2 12 2nithika 2 2 7.5 1 8 1vignesh 12 1 31.5 4 38 4ayyapan 4 1 12.9 2 14 2sandhya 3.5 2 13.7 2 14 2srimathy 7 2 21.9 3 26 3nagoremee 8 2 20.3 3 24 3navin 11 1 19.8 2 22 3arulsandha 5 1 11.3 2 12 2kani 3 2 11.4 2 12 2anbu 4.5 1 13.2 2 14 2kumanan 12 1 26 3 30 3guru 10 1 29 3 32 4karthick 2.5 1 10.3 2 12 2sudha 7 2 20 2 24 3vanitha 4 2 13 2 14 2dhuruvan 8 1 22.6 3 28 3geetha 4.5 2 15.1 2 16 2kumaresan 6.5 1 19 2 22 3kannapan 7.5 1 23.7 3 28 3anbukarasi 3.5 2 14 2 16 2ilakia 6 2 16.6 2 18 2kannan 5 1 15.7 2 18 2ramesh kum 11 1 27 3 30 3krithick 9 1 25.6 3 28 3keerthi 3 2 12 2 14 2arulprakash 2.5 1 10.5 2 12 2aravind 3 1 10 1 12 2soundar 3 1 13.3 2 14 2vinupriya 12 2 25.4 3 28 3arjunprabh 4 1 16 2 18 2dhanusri 7 2 20.3 3 22 3varun 7 1 21.2 3 22 3chandru 10 1 26 3 32 4

snehan 2 1 12 2 12 2aarav 3.5 1 15 2 14 2anupama 2 2 11 1 12 2sherly 1 2 9 1 10 1princy 5 2 20 2 24 3mary 5 2 18 2 20 2mariam 7 2 22 3 24 3preethi 6 2 19 2 24 3julie 9 2 27 3 32 4anbu 8 1 30 3 34 4rajaram 7 1 25 3 28 3prabha 5.5 1 22 3 24 3shalini 1.5 2 10 1 12 2shailaja 1 2 8 1 10 1kulandai 3 1 13 2 12 2rambabu 2 1 12 2 12 2paneer 8 1 24 3 30 3nalan 1 1 7.5 1 8 1nimisha 0.4 2 4.8 1 4 1manisha 1.5 2 10.5 2 12 2bharath 2 1 13 2 14 2velammal 2 2 13.5 2 14 2kathir 0.5 1 5.5 1 6 2chathriyan 1 1 8 1 10 1kesav 5 1 20 2 24 3gourav 4 1 16 2 18 2keerthy 9.5 2 32 4 40 4williams 2.5 1 14 2 16 2koushik 5.5 1 20 2 24 3vivek 6.6 1 28 3 32 4sabarivasan 5.5 1 24 3 26 3garudan 4.5 1 18 2 22 3dhanush 3.5 1 14 2 16 2dev 2.5 1 13 2 14 2diya 9 2 31 4 36 4dhruvan 8 1 28 3 34 4kamala 7 1 28 3 34 4kamal 6 1 24 3 26 3devan 5 1 22 3 24 3ponraj 4 1 17 2 20 2ponpaandi 3 1 14 2 16 2stallin 2 1 11 2 12 2periyar 10 1 35 4 40 4hitler 11 1 32 4 38 4jackie 0.2 1 4 1 4 1maaran 1 1 7.5 1 8 1prasanth 0.5 1 5 1 6 1karupan 2.5 1 13 2 14 2muthu 2.5 1 11.5 2 12 2meera 2 2 12 2 12 2

nesamani 9 1 30 3 38 4christophe 7 1 28 3 32 4varathan 3.5 1 16 2 20 2dheivanai 3 2 13 2 14 2muthamil 8 2 28 3 32 4kaleel 7.5 1 30 3 34 4ibrahim 5.5 1 21 3 24 3syed aham 5 1 24 3 28 3vetrivel 0.6 1 7 1 8 1velusamy 0.2 1 4 1 4 1velmuruga 0.5 1 5.5 1 6 1velunaiker 1.5 1 9 1 10 1bobby 1 1 8.5 1 10 1sarath 1 1 8 1 10 1babuji 4 1 17 2 18 2samrutha 2 2 12 2 14 2amsavarthi 9 2 35 4 40 4malini 5.5 2 23 3 26 4oviya 4 2 18 2 20 2ramky 7.5 1 29 3 34 4arjun 7 1 31 4 34 4arasan 5 1 20 2 22 3daniel 1.5 1 10 1 12 2dhilip 0.5 1 5.5 1 6 1justin 0.5 1 5 1 6 1joel 1.5 1 8.9 1 10 1jacob 2.5 1 12.5 2 14 2nestle 3.5 1 15.5 2 14 2arockiyam 3 1 13 2 16 2manju 8 2 29 3 34 4arushi 0.5 2 6.3 1 8 1westley 4.5 1 20 2 24 3sivaraman 2.5 1 13 2 14 2venkat 11 1 34 4 40 4abubacker 9 1 30 3 36 4abi 0.5 2 5 1 6 1sheik 6.5 1 23 3 24 3rupesh 4 1 16 2 18 2natraj 4 1 18 2 22 3mathi 3.5 1 14 2 16 2raju 2 1 13 2 14 2murugesh 10 1 35 4 40 4varnesh 0.5 1 5.8 1 6 1sarath 7 1 26 3 32 4boopathy 4.6 1 19 2 24 3boopalan 3.5 1 13.7 2 14 2nimilan 5 1 17.4 2 20 2pinky 10 2 41 4 38 4kiruba 12 1 37.8 4 42 4devaki 3 2 12.9 2 14 2

kannan 4 1 16.9 2 20 2dharman 1 1 7.5 1 8 1murugesh 3 1 13 2 14 2mayil 7 1 24 3 26 3kumar 9 1 34 4 38 4revathy 8 2 29 3 34 4revathi 5 2 19 2 22 3ramprabhu 4 1 16 2 16 2vicky 7 1 28 3 34 4logesh 3 1 13 2 14 2logeshwara 6.5 1 21 3 22 3pranesh 2 1 12.5 2 14 2suresh 1.5 1 10 1 10 1valavan 12 1 45 4 38 4raghuvaran 11 1 30 4 36 4guru 10 1 28 3 34 4rahavan 9 1 35 4 34 4prabhu 8 1 32 4 28 3karthikeyan 7 1 30 3 34 4vikram 6 1 26 3 28 3dhruva 5 1 20 2 22 3santhosh 4 1 16 2 18 2chandru 4.5 1 15 2 16 2chandran 3 1 12.5 2 14 2kumari 3.5 2 15 2 14 2kaanchana 2.5 2 13 2 14 2cinthiya 2 2 12 2 14 2sindhu 1.5 2 11.7 2 12 2sindhuja 1 2 8.9 1 10 1vanthana 0.5 2 4.9 1 6 1vatchala 0.5 2 5.3 1 6 1varuni 2 2 11.6 2 12 2dhivya 3.5 2 14.2 2 14 2manohari 5 2 17.3 2 16 2sudharsan 3 1 14.1 2 14 2velmuruga 2.5 1 13.9 2 14 2palani 3.5 1 13.9 2 14 2vijay 3 1 14.5 2 14 2sethu 1 1 8.3 1 10 1karupan 6 1 29 3 34 4goutham 7 1 27.3 3 26 3sabari 3 1 14.1 2 14 2visu 8 1 32 4 38 4jakan 9 1 37 4 44 4kesavan 2 1 12.8 2 14 2pethukani 1 2 7.9 1 10 1dorian 1.5 1 9.2 1 10 1selvakuma 3 1 13.9 2 14 2selvaraj 6 1 21.9 3 26 3kamesh 4 1 15.8 2 18 2

rajkumar 8 1 34.8 4 42 4balaji 9 1 40.7 4 36 4bala 10 1 34.7 4 40 4balan 12 1 32.9 4 44 4ganapathy 2.5 1 13.6 2 14 2vaani 4.5 2 17.2 2 18 2laxmi 6.5 2 21.6 3 24 3laxmanan 8 1 28 3 34 4sasivarnan 7 1 28 3 32 4vetrivel 2 1 12.8 2 14 2yashvanth 1.5 1 11.8 2 12 2nithish 0.5 1 5.2 1 6 1thulasi 12 2 46 4 36 4kanika 11 2 34 4 40 4kanishka 3 2 12.9 2 14 2anisha 2.5 2 12.7 2 14 2anusha 5 2 18.9 2 22 3arun 6 1 24.5 3 22 3pradeep 4 1 16 2 18 2mano 8 1 30 3 28 3lalitha 9 2 29.5 3 34 4kiran 12 2 28 3 40 4suresh 4.5 1 15.9 2 18 2vinoth 4.5 1 17 2 20 2karthikeyan 6 1 19.2 2 24 3mahesh 7 1 25.2 3 28 3madhan 8 1 28 3 34 4viji 9 2 29 3 38 4kesavan 2 1 12.5 2 14 2murali 1 1 8.4 1 10 1pushkar 1.5 1 10 1 12 2pooja 1.5 2 9.9 1 10 1kirthi 1 2 8.5 1 10 1keerthana 0.5 2 6 1 6 1nandhini 1 2 7.9 1 10 1ilakia 1 2 9 1 10 1kavitha 2 2 12.3 2 14 2vanitha 2.5 2 13.1 2 14 2rajkamal 3.5 1 13.9 2 14 2rajesh 4.5 1 15.7 2 16 2arul 5.5 1 16.3 2 16 2paandi 6 1 21 3 22 3soundar 7 1 26.2 3 30 3vijay 6.5 1 27.4 3 30 3anurega 3 2 13.3 2 14 2aravind 5 1 19.1 2 22 3sumathi 8 2 31 4 38 4karuppu 9 1 29 3 34 4muthu 9.5 1 41 4 28 3b/o rani 5 2 20 2 22 3

rahul 6 1 23.2 3 26 3krishnan 2.5 1 12 2 12 2arun 3 1 15 2 14 2shasthika 1 2 7 1 8 1kiruba 3 1 11 2 10 1kirthi 7 2 17 2 16 2kannan 10 1 28 3 30 3niranjan 5.5 1 16 2 17 2mani 0.5 1 5 1 6 1manikanda 9 1 25 3 24 3jakan 12 1 30 3 32 4janani 10 2 32 4 36 4sowmiya 8 2 29 3 32 4sanjana 2 2 10 1 12 2saran 2.5 1 11 2 12 2aaradhna 2 2 9 1 10 1kamal 1 1 7.5 1 8 1ajish 1.5 1 8.5 1 8 1lalitha 7 2 20 2 20 2kiran 6 1 23 3 22 3suresh 5 1 20 2 20 2vinoth 5 1 21 3 22 3karthikeyan 5 1 27 3 30 3mahesh 9 1 30 3 34 4madhan 8 1 28 3 32 4viji 10 2 35 4 38 4kesavan 11 1 39 4 38 4murali 6 1 23 3 22 3pushkar 4.5 1 20 2 22 3pooja 8 2 30 3 36 4kirthi 8.8 2 28 3 34 4keerthana 9 2 31 4 36 4nandhini 9 2 33 4 38 4ilakia 11 2 32 4 38 4kavitha 2 2 11 2 12 2vanitha 1.5 2 9 1 10 1rajkamal 3.5 1 14 2 14 2rajesh 6 1 20 2 20 2arul 8 1 30 3 26 3paandi 7 1 28 3 30 3soundar 4 1 15 2 14 2vijay 3.5 1 13 2 12 2anurega 4 2 14 2 14 2aravind 2.5 1 11 2 12 2sumathi 6 2 18 2 16 2karuppu 8 1 25 3 24 3muthu 4 1 13 2 12 2b/o rani 0.1 2 4 1 4 1rahul 5 1 17 2 16 2sudharsan 4.7 1 15 2 14 2

gokul 6.2 1 20 2 18 2vasanth 6.7 1 21 3 20 2thomas 7 1 24 3 26 3nilofer 8 2 27 3 28 3asmath 7.5 1 25 3 26 3asmaas 7 1 26 3 30 3amul 3 1 13 2 12 2raja 7 1 20 2 20 2rajashri 9 1 27 3 30 3vasanthi 10 2 31 4 36 4kalpana 11 2 33 4 38 4suganya 6 2 20 2 18 2anitha 3 2 11 2 12 2kannan 2 1 10 1 10 1karthikeyan 5 1 15 2 14 2rani 2 2 11 2 12 2durai 2.5 1 14 2 12 2gowtham 8 1 23 3 26 3shoban 6 1 20 2 24 3harish 2 1 12 2 10 1nithish 1.5 1 9 1 10 1nirupama 1 2 8 1 8 1claret 1 1 8.5 1 8 1anton 3 1 13 2 12 2suganthi 3 2 13.5 2 12 2suthan 3.5 1 15 2 14 2lohith 7 1 20 2 18 2sujith 5 1 18 2 18 2aathi 3 1 14 2 14 2nagul 2 1 11 2 12 2vijay 5 1 18 2 18 2ajith 9 1 31 4 36 4setupathi 2.5 1 12 2 12 2kumaran 1.5 1 9 1 10 1jana 0.5 1 4.5 1 6 1balaji 2 1 11 2 10 1arunan 4 1 17 2 16 2valli 5 2 16 2 16 2petchi 6 2 20 2 18 2karuppan 7 1 21 3 24 3kanmani 10 2 35 4 40 4kishore 2 1 11 2 12 2vignesh 5 1 19 2 18 2varathan 1 1 9 1 10 1senthil 0.3 1 6 1 6 1kanthan 3 1 13 2 12 2vishal 0.5 1 5.5 1 6 1surya 7 1 20 2 18 2shtuthi 4 2 15 2 14 2venmathi 8 2 22 3 20 2

girija 2 2 12 2 12 2suganya 4 2 16 2 16 2thangam 9 2 30 3 38 4veena 11 2 29 3 36 4vanthana 6 2 21 3 20 2selvam 4.5 1 18 2 16 2selvi 6.5 2 25 3 24 3rajgopal 3 1 12 2 12 2ramanan 1 1 8 1 10 1rajiv 3 1 13 2 12 2anbu 2.5 1 14 2 14 2rajaram 6 1 24 3 22 3joseph 7 1 23 3 24 3john 5 1 20 2 24 3christy 1 1 8 1 10 1jameela 1.5 2 9 1 8 1mohd.sam 3 1 12 2 10 1kiruba 3.5 2 12 2 10 1sameera 6 2 23 3 22 3diana 9 2 28 3 34 4daisy 2 2 11 2 10 1ganesh 3.5 1 15 2 14 2perumal 1 1 8 1 8 1ramu 1 1 7.7 1 10 2rasiga 1.5 2 9.4 1 8 1santhi 3 2 13.6 2 16 2jeyanthi 4 2 15 2 14 2jaya 6 2 18 2 22 3karnan 5.5 1 17 2 18 2gopal 5 1 17 2 16 2harish 2 1 11 2 10 1jagadesh 2.5 1 13 2 12 2govarthan 6 1 18 2 16 2girija 5.5 2 17 2 14 2girish 4 1 16 2 16 2saran 1 1 9 1 10 2sarath 1.3 1 10 1 10 2saratha 8 2 28 3 34 4saravanan 10 1 30 3 34 4tamizh 4 1 17 2 16 2nagulan 3 1 15 2 14 2surendar 4 1 16 2 14 2sarathy 5.5 1 18 2 20 2priya 8 2 26 3 34 4priya 2 2 12 2 12 2nisha 1 2 7.9 1 8 1jeslin 3 2 14.7 2 12 2bakrudhen 12 1 36 4 40 4jayakumar 2 1 13 2 12 2rajavel 6 1 19 2 20 2

kumar 3 1 14 2 14 2vennila 2 2 11 2 12 2varun 1.5 1 10 1 10 2bharathi 7 2 22 3 22 3ilamaran 4 1 18 2 16 2seeman 9 1 32 4 36 4ramasamy 2 1 12 2 12 2rajathi 2.5 2 13 2 12 2rajan 6 1 25 3 22 3vaishnavi 4 2 19 2 26 3vasikaran 3.5 1 15 2 14 2vaishali 1 2 9 1 8 1vaijayanthi 0.5 2 6 1 8 1karthiga 10 2 Feb-00 4 40 4kirthika 7 2 27 3 32 4prasanna 8 1 26 3 30 3prasath 1 1 8.5 1 8 1prabakar 1.5 1 9.5 1 10 1pranesh 1 1 8 1 10 1prema 1 2 9 1 10 1pradeep 0.5 1 5.5 1 6 1sankar 11 1 33 4 36 4ramya 12 2 38 4 40 4ramesh 2 1 11 2 12 2rajini 3 1 14 2 14 2siva 5 1 20 2 20 2nagarajan 6 1 24 3 22 3sundar 8 1 29 3 36 4sudhir 2 1 11 2 12 2arunan 1 1 9 1 10 1aruna 1.5 2 11 2 12 2rajeev 3.5 1 16 2 16 2pachamuth 4.5 1 18 2 20 2sivnesh 2.5 1 14 2 14 2sivakaran 3 1 17 2 18 2muzzamil 2 1 12 2 12 2sanjay 1 1 9 1 10 1kicha 1.5 1 11 2 12 2sham 3 1 15 2 16 2harini 6 2 21 3 24 3paapu 8 2 30 3 36 4padma 9 2 32 4 38 4lacanya 3 2 12 2 12 2latchana 5 2 17 2 20 2lagubaran 7 1 21 3 26 3mano 10 1 33 4 38 4manoj 2 1 12 2 12 2manoharan 2 1 12 2 12 2pallavan 1 1 9 1 10 1harshini 3 2 14 2 16 2

jothi 2 2 12 2 12 2vetrivel 2.5 1 12 2 14 2yashvanth 1.5 1 10 1 10 1nithish 6 1 22 3 24 3thulasi 1.5 2 10 1 12 2anushri 1 2 8 1 10 1rithesh 1.5 1 11 2 12 2arun 1.5 1 11 2 12 2peadeep 2.5 1 14 2 12 2devandran 2 1 10 1 12 2pandi 4 1 15 2 16 2shanmugam 7 1 20 2 22 3naveen 2.5 1 14 2 16 2abdul 5 1 22 3 22 3arivarasan 8 1 30 4 34 4siva 12 1 36 4 40 4dharsini 0.5 2 5.5 1 6 1tarun 4 1 16 2 16 2bathrakaal 4.5 2 17 2 16 2ashwin 3 1 11 2 12 2vennila 2.5 2 12 2 12 2kanmani 8 2 29 3 34 4yaashrai 6 1 25 3 28 3sanjana 4.5 2 20 2 22 3rishwana 8 2 28 3 34 4martin 9 1 30 3 34 4vimal 1 1 9 1 10 1sasi 2 1 12 2 12 2dinesh 12 1 35 4 40 4swetha 2.5 2 15 2 16 2sham 5 1 19 2 17 2lingam 11 1 32 4 36 4vicky 5 1 20 2 22 3rajpriya 7.5 2 28 3 34 4ashwin 6 1 28 3 36 4ramani 0.5 2 5 1 6 1john 9 1 29 3 34 4suber 12 1 38 4 42 4b/o ramya 0.2 2 4 1 4 1mumtaz 11 2 31 4 34 4sanish 2 1 12 2 12 2kirmani 1 1 9 1 10 1dhoni 11 1 32 4 42 4dharman 3 1 14 2 12 2murugesh 4 1 17 2 14 2mayil 8 1 29 3 36 4kumar 5 1 20 2 22 3revathy 0.7 2 7.5 1 10 1revathi 2 2 12.5 2 12 2ramprabhu 1.5 1 11 2 12 2

vicky 1.5 1 10.5 2 12 2logesh 2.5 1 13 2 12 2logeshwara 3 1 12.5 2 12 2pranesh 4 1 15 2 14 2suresh 8 1 27.5 3 32 4valavan 9 1 30 3 32 4raghuvaran 1 1 8.5 1 10 1guru 0.4 1 5 1 6 1rahavan 4 1 16 2 16 2prabhu 3 1 14 2 14 2karthikeyan 7 1 24 3 28 3vikram 9 1 29 3 34 4dhruva 8 1 28 3 32 4santhosh 5 1 20 2 22 3chandru 4 1 17 2 20 2chandran 7 1 26 3 30 3kumari 3 2 14 2 15 2kaanchana 6.5 2 24 3 28 3cinthiya 2 2 11 2 12 2sindhu 1.5 2 10.5 2 12 2sindhuja 12 2 36 4 42 4vanthana 11 2 35 4 42 4vatchala 10 2 35 4 42 4varuni 9 2 32 4 42 4dhivya 8 2 27 3 34 4manohari 7 2 28 3 34 4sudharsan 6 1 25 3 30 3velmuruga 5 1 20 2 22 3palani 4 1 17.6 2 20 2vijay 4.5 1 22 3 24 3sethu 3 1 16 2 20 2karupan 3.5 1 15 2 18 2goutham 2.5 1 12 2 12 2sabari 2 1 10.5 2 12 2visu 1.5 1 11 2 12 2jakan 1 1 8.5 1 10 1kesavan 0.5 1 5.5 1 6 1pethukani 0.5 2 5.5 1 6 1dorian 2 1 11 2 12 2selvakuma 3.5 1 13 2 14 2selvaraj 5 1 18 2 20 2kamesh 3 1 13 2 15 2rajkumar 2.5 1 12 2 14 2balaji 3.5 1 12 2 14 2bala 3 1 14 2 12 2balan 1 1 8 1 10 1ganapathy 6 1 21 3 20 2vaani 7 2 25 3 20 2laxmi 3 2 13 2 12 2laxmanan 8 1 30 3 26 3

sasivarnan 9 1 27 3 24 3sanjay 1 1 7.2 1 7 1ramesh 0.5 1 5.5 1 6 1praksh 0.2 1 4 1 4 1parnetha 2 2 12 2 12 2kajal 3.5 2 14.2 2 14 2kajol 6 2 19.2 2 20 2raveena 5 2 21 3 20 2dhilip 4.5 1 17.1 2 18 2sunny deol 7 1 25 3 24 3shrukh 9 1 31 4 36 4salman 11 1 30 3 38 4aamir 3.5 1 14.1 2 14 2aaradhna 0.3 2 4.7 1 4 1sanjeev 1 1 8.2 1 10 1anitha 1.5 2 9.9 1 10 1avila 4 2 15.9 2 16 2ananya 7 2 22.8 3 26 3ananyasri 3 2 12.9 2 14 2saritha 2.5 2 13 2 14 2sangeetha 12 2 39 4 28 3sangitha 2.4 2 10.9 2 12 2sangeeth 1.5 1 9.7 1 10 1maarimuth 3 1 12.5 2 12 2aishwarya 6 2 20.5 3 22 3aniruth 4 1 15 2 16 2iman 3.5 1 12.9 2 14 2vadivel 0.4 1 4.5 1 6 1singamuthu 1 1 7.5 1 8 1saravanan 2 1 11.9 2 12 2uday 3 1 13.5 2 14 2chandar 2.5 1 13 2 14 2chandru 2 1 12.1 2 12 2charlie 1 1 7.9 1 8 1salim 0.5 1 5.3 1 6 1karnan 0.6 1 6 1 6 1lolitha 2.5 2 13.1 2 14 2harsha 5 2 19 2 22 3robin 6 1 19.9 2 22 3davidson 7.5 1 25 3 28 3harrison 2.5 1 13.5 2 14 2nelson 1 1 9 1 10 1vijay 1 1 8.6 1 10 1anoop 1.7 1 10.2 2 12 2piyush 0.4 1 5.1 1 6 1gupta 10 1 31 4 24 3ganga 11 2 29 3 40 4komal 3 2 12.9 2 14 2valli 4 2 14 2 16 2brindha 5 2 19.5 2 22 3

riya 2 2 12.4 2 14 2roopa 3 2 13.7 2 14 2roobini 5 2 21 3 26 3idhaya 8 2 29 3 34 4rohini 6 2 25 3 28 3roshini 2.5 2 12 2 12 2vasantha 1.5 2 10.9 2 12 2meena 3 2 12.7 2 14 2nirosha 2 2 11.9 2 12 2raahini 2.5 2 13.1 2 14 2raadhika 1.5 2 10 1 12 2varsha 1 2 9.1 1 10 1venamathi 2.5 2 13 2 14 2megha 5 2 20.1 3 22 3yazhini 6 2 18.3 2 20 2shamili 7 2 20.5 3 24 3sharvani 2 2 10.9 2 12 2kalyani 2.5 2 11.2 2 12 2kannan 1.5 1 9.7 1 12 2siddharth 0.4 1 5.9 1 6 1vasanth 1 1 8.9 1 10 1vedha 1.5 1 9.8 1 12 2vickram 2 1 10.8 2 12 2shri 3 1 13.5 2 16 2kaniyan 2.5 1 12.9 2 14 2subbu 4 1 14.2 2 16 2poongundr 3 1 13.5 2 14 2medley 3.5 1 13 2 14 2kaliyugan 5 1 17.5 2 20 2bramman 6.5 1 19.2 2 22 3nelson 10 1 30 3 36 4gahi 11 1 29 3 34 4varunan 2.5 1 12.6 2 14 2sarav 1.5 1 11 2 12 2kicha 1 1 9 1 10 1dhaksin 0.5 1 5.2 1 6 1charles 1 1 7.9 1 10 1claret 1 1 8.3 1 10 1rooban 1 1 8.7 1 10 1dharsan 1.5 1 9.3 1 12 2boopalan 2 1 11.4 2 14 2jakan 2.5 1 12.9 2 14 2kesavan 3 1 13 2 16 2nayagan 3.5 1 13.4 2 16 2soori 3 1 13 2 16 2vadivel 4 1 14.2 2 16 2vivek 4.5 1 13.9 2 16 2prem 2.5 1 12 2 14 2karthikeyan 2 1 11.2 2 14 2kennedy 1 1 9 1 10 1

prabhu 1 1 8.9 1 10 1

ANNEXURE V- CONSENT FORM IN REGIONAL LANGUAGE (TAMIL)

ஆரா��சி ஒ��த� ப�வ�

ேததி:

ஆரா��சி தைல��:

ெபய�:

வய�: உ�ேநாயாள� எ�:

பா�: ஆரா��சி ேச��ைக எ�:

இ�த ஆரா��சிய�� வ�பர�க�� அத� ேநா�க�க�� ��ைமயாக என��

வ�ள�க�ப�ட�. என�� வ�ள�க�ப�ட வ�ஷய�கைள ����ெகா�� என� ச�மத�ைத

ெத�வ��கிேற�.

இ�த ஆரா��சிய�� ப�ற�� நி�ப�தமி�றி எ� ெசா�த வ���ப�தி� ேப�� நா�

ப�� ெப�கிேற�, ம��� நா� இ�த ஆரா��சிய�லி��� எ�த ேநர�� ப��வா�கலா�

எ�பைத��, அதனா� எ�த பாதி��� ஏ�படா� எ�பைத��, நா� ���� ெகா�ேட�.

நா� எ��ைடய �யநிைன�ட� ம��� �� �த�திர��ட� இ�த ம���வ

ஆரா��சிய�� எ�ைன ேச����ெகா�ள ச�மதி�கிேற� ���கைள அ�ல�

க����கைள ெவள�ய��� ேபாேதா அ�ல� ஆரா��சிய�� ேபாேதா எ��ைடய

ெபயைரேயா அ�ல� அைடயாள�கைளேயா ெவள�ய�டமா�டா�க� எ�பைத�� அறி��

ெகா�ேட�.

என� ேநாய�� த�ைம ம��� ப��வ�ைள�கைள�� ��ைமயாக

����ெகா�ேட�. இ�த ஆரா��சிய�� என� ேநாய�� �ல��� ம��� த�ைமைய

ம��ேம ஆரா�வா�க� எ�பைத அறி�� ெகா�ேட�.

இதனா� எ� ைவ�திய �ைறகள�� எ�த மா�ற�� பா�ைவ�திறன�� எ�தவ�த

பாதி��� ஏ�படா� எ�பைத�� ெத���ெகா�ேட�. என�� வ�ள�க�ப�ட

வ�ஷய�கைள ��ைமயாக ����ெகா�� இ�த ஆரா��சிய�� ப��ெகா�ள எ� ��

மன�ட� ஒ���ெகா�கிேற�.

ஆரா��சியாள� ைகெயா�ப� ப�ேக�பாள� ைகெயா�ப�

ேததி:

Urkund Analysis Result Analysed Document: Reliability of broselow tape-length based emergency weight

estimation.docx (D31688018)Submitted: 10/25/2017 9:48:00 PM Submitted By: [email protected] Significance: 0 %

Sources included in the report:

Instances where selected sources appear:

0

U R K N DU

CERTIFICATE

This is to certify that this dissertation work titled “RELIABILITY OF

BROSELOW TAPE-LENGTH BASED EMERGENCY ESTIMATION IN

PAEDIATRIC POPULATION OF SOUTHERN TAMIL NADU” of the

candidate Dr.D.SANJEEV KUMAR with registration number 201517105 for the

done for award of M.D Degree Branch VII – PAEDIATRIC MEDICINE. I

personally verified the urkund.com website for the purpose of plagiarism check. I

found that the uploaded thesis file contained from introduction to conclusion pages

and result shows 0% of plagiarism in the dissertation.

Guide and Supervisor sign with seal