Measuring Skilled Attendance in the uThungulu District 2 Mar 2

7/27/2019 Measuring Skilled Attendance in the uThungulu District 2 Mar 2

1/129

Measuring Skilled Attendance in the uThungulu

District, KwaZulu-Natal, in 2008

A dissertation submitted to the

Department ! "u#lic $ealth Medicine

Nelsn %& Mandela Schl ! Medicine

'ni(ersit) ! KwaZulu-Natal

Dur#an, Suth A!rica

*n partial !ul!ilment ! the re+uirements !r the

Master in "u#lic $ealth

)

Slange Mianda

Super(isr Dr Anna .ce

March 20/0


2/129

As the candidate1s super(isr * agreed nt agree t the su#missin ! this

dissertatin3

Super(isr 444444444& Date 4444444444&


3/129

III

Dedication

I dedicate this work to my Lord Jesus Christ author and finisher of my faith, author of life, my

everything. Thank you for the gift of time and life.

To my family, thank you for your love, support and belief in me.


4/129

I

Acknowledgements

I would like to thank!

The "niversity of #wa$ulu%&atal 'cholarship Committee for supporting my studies throughout

the years

The #wa$ulu%&atal (rovincial )epartment of *ealth +esearch ffice and the uThungulu *ealth

)istrict for granting permission to conduct this study

All hospital managers in the study hospitals -Catherine ooth, /kombe, /showe, &kandla and 't

0ary1s hospitals2 for supporting this study

All midwives for taking part in the obstetric knowledge and skills test

The )epartment of (ublic *ealth 0edicine

(rof Anna Coutsoudis for introducing me to the university community

(rof /ddie 0hlanga -)epartment of bstetrics and 3ynaecology2, (rof *ugh (hilpott and 0s

$o 04olo -Centre for +ural *ealth2 for your input toward the implementation of this study

)r Anna oce, my supervisor, for showing interest in the study and for providing constructive

advice and active support. *er patience and encouragement have helped me achieve more than I

thought I could

All my 0(* colleagues

5riends at 3lenridge Church International, ridget Thomas, /mma Anyachebelu, ukky

Ade6umo, Celdi Lu4olo for your friendship and spiritual support.


5/129

Declaration

I 'olange 0ianda declare that!

-i2 The research reported in this dissertation, e7cept where otherwise indicated, is my original

work.

-ii2 This dissertation has not been submitted for any degree or e7amination at any other

university.

-iii2 This dissertation does not contain other person1s data, pictures, graphs or other

information, unless specifically acknowledged as being sourced from other persons.

-iv2 This dissertation does not contain other persons1 writing, unless specifically acknowledged

as being sourced from other researchers. 8here other written sources have been 9uoted!

a2 Their words have been re%written but the general information attributed to them has been

referenced.

b2 8here their e7act words have been used, their writing has been placed inside 9uotation

marks, and referenced.

-v2 8here I have reproduced a publication of which I am an author, co%author or editor, I have

indicated in detail which part of the publication was actually written by myself alone and have

fully referenced such publications.

-vi2 This dissertation does not contain te7t, graphics or tables copied and pasted from the

internet, unless specifically acknowledged, and the source being detailed in the dissertation and

the references section.

Signature 444444444& Date 4444444444&


6/129

I

Abstract

ackgrund

The 0illennium )evelopment 3oals call for two%third and three%9uarter reductions in (erinatal

0ortality +ates and 0aternal 0ortality +atios. The main strategy towards achieving these

reductions is to increase access to skilled attendance. *owever, it cannot be confirmed that all

health professionals are skilled in managing women in labour, nor that they are functioning in

enabling environments. To measure the provision of skilled attendance, this study was

undertaken in five Level : *ospitals in the uThungulu *ealth )istrict of #wa$ulu%&atal. The

ob6ectives of the study were!

:. To establish perinatal outcomes for each Level : *ospital in uThungulu *ealth

)istrict.

;. To evaluate the 9uality of intrapartum care provided in Level : *ospitals in

uThungulu *ealth )istrict.

. To evaluate the environment in which births are attended in Level : *ospitals in


?. Compare the 9uality of care, the knowledge, skills and environment with perinatal

outcomes.

Methds(erinatal outcomes -(&0+, 5'+, /&&)+ and (CI2 were calculated for each hospital@

maternity case records of women who have delivered in these Level : *ospitals were audited

to assess the 9uality of intrapartum care@ obstetric knowledge and skills of midwives were

assessed@ as was the enabling environment within which midwives worked, which included a

measurement of their workload. A correlation between perinatal outcomes and the 9uality of

intrapartum care, knowledge and skills and the enabling environment was performed to

determine whether variables were associated.


7/129

II

%esults

The overall (&0+ for five hospitals in uThungulu *ealth )istrict was


8/129

III

Table of Contents

)/)ICATI&........................................................................................................................IIIAC#&8L/)3/0/&T'........................................................................................................I

)/CLA+ATI&......................................................................................................................A'T+ACT...........................................................................................................................IT*+//*'(ITAL'0/TT*//&ALI&3/&I+&0/&T'TA&)A+). ALL*'(ITAL'"T&/

'C+/)(+LF&+/5/++AL, A&)T*/AAILAILITF5'"(/+I'I&&T*'*I5T'. &/*'(ITAL'C+/)(+LF&)+"3'A&)'"((LI/'. /+ALL&*'(ITAL'+/(+T/)T*/(+/'/&C/5ALLT*//L/0/&T'5T*//&ALI&3/&I+&0/&T. T*+//*'(ITAL'*A)ACC/(TAL/8+#LA)'. .....................................................................................................II

TAL/5C&T/&T'........................................................................................................IIILI'T5TAL/'...................................................................................................................GI)/5I&ITI&5T/+0'......................................................................................................GIIIAC+&F0'.........................................................................................................................G

LI'T5A+/IATI&'...................................................................................................GI:. AC#3+"&)...................................................................................................................:

:.: I&T+)"CTI&............................................................................................................::.; (+L/0'TAT/0/&T................................................................................................>.; H"ALITF5I&T+A(A+T"0CA+/...........................................................................>>.> T*//&ALI&3/&I+&0/&T.................................................................................?;

T*+//*'(ITAL'0/TT*//&ALI&3/&I+&0/&T'TA&)A+). ALL*'(ITAL'"T&/'C+/)(+LF&+/5/++AL, A&)T*/AAILAILITF5'"(/+I'I&&T*'*I5T'. &/*'(ITAL'C+/)(+LF&)+"3'A&)'"((LI/'. /+ALL&*'(ITAL'+/(+T/)T*/(+/'/&C/5ALLT*//L/0/&T'5T*//&ALI&3/&I+&0/&T. 'T")I/'F3A&3)A)//TAL

-;


11/129

GI

List of Tables

List of Figures

5I3"+/=. :! 0/A&/+ALL(/+C/&TA3/'C+/(/++/C+), FL//L: *'(ITALI&"T*"&3"L"*/ALT*)I'T+ICT, J"&/;D............................................................................=

5I3"+/=.

5I3"+/=. =! LA"+3+A(*/G/+CI'/II O 0/)IA&(/+C/&TA3/'C+/FL//L: *'(ITALI&"T*"&3"L"*/ALT*)I'T+ICT, J"&/;D........................................................................=

5I3"+/=. >! ((* 'TATI&% /+ALL0/)IA&(/+C/&TA3/'C+/, FL//L: *'(ITALI&"T*"&3"L"*/ALT*)I'T+ICT, J"&/;D............................................................................>:


12/129

GII

List of Appendices

A((/&)IG:. :! ('T3+A)"AT//)"CATI&C00ITT//A((+AL.............................?A((/&)IG:. ;! I0/)ICAL+/'/A+C*/T*IC'C00ITT//A((+AL........................

A((/&)IG:.

A((/&)IG:. =! *'(ITAL(/+0I''I&P'"((+TL/TT/+'................................................A((/&)IG:. >! L/TT/+T0&3L8A&/*'(ITAL....................................................D=A((/&)IG:. ?! I&5+0/)C&'/&T5+0.......................................................................D>

A((/&)IG:. :! ('T3+A)"AT//)"CATI&C00ITT//A((+AL.............................?A((/&)IG:. ;! I0/)ICAL+/'/A+C*/T*IC'C00ITT//A((+AL........................A((/&)IG:.


13/129

GIII

Definition of Terms

/arly neonatal death rate The number of live born infants that die in the first week of life per

: live born infants -(/( ;D, p=2.

Level : *ospital A health facility that provides basic obstetric, surgical, medical,paediatric and psychiatric care. Includes anaesthetic facility -oce;>, p7ii2.

0aternal mortality ratio A maternal death is the death of a woman while pregnant orwithin =; days of the termination of pregnancy, irrespective of theduration and site of the pregnancy, from any cause related to or

aggravated by the pregnancy or its management but not fromincidental or accidental causesM -8* :;, p=2. The maternalmortality ratio is the number of maternal deaths in a given time

period per : live births during the same time period-Abou$ahr and 8ardlaw ;=, p=2.

0idwife A person who has successfully completed the prescribed course ofstudies in midwifery and ac9uired the re9uisite 9ualifications to beregistered andPor legally licensed to practice midwifery-8*PIC0P5I3 ;=, p2.

(artograph -partogram2 The partograph is a simple chart to record and monitor the progressof labour and other essential maternal and fetal observations. It can

provide an early indication that emergency obstetric care is needed-"&5(A ;=a, p:2.

(erinatal inde7 care The ratio of the perinatal mortality rate to the low birth weight rate-(/( ;D, p>2.

(erinatal mortality rate &umber of all perinatal deaths per : total births. (erinataldeaths include all stillbirths that weigh more than > grams andearly neonatal deaths -(/( ;D, p=2.

'killed attendant *ealth professional providing care during childbirth-8*PIC0P5I3 ;=, p2. This professional needs to have aset of skills, defined as core midwifery skills, to provide effectivecare during childbirth. A skilled attendant needs to be able to


14/129

GI

conduct normal deliveries and recogni4e, manage and referobstetric complications -8*P IC0P5I3 ;=, p2.

'killed attendance The process by which women are provided with ade9uate careduring labour, delivery and the early postpartum period and

re9uires both skilled attendant and the enabling environment-8*PIC0P5I3 ;=, p2.

'tillbirth +efers to an infant born dead after ? months of intra%uterine life-(/( ;D, p=


15/129

G

Acronyms

)/(A0 )/C/&T+ALI'/)(+3+A00/5+A)A&C/)0I)8I/'

5I3 International 5ederation of 3ynaecology and bstetrics

IC0 International Confederation of 0idwives

0)3 0illennium )evelopment 3oals

+C3 + FALCLL/3/5'T/T+ICI&A'A&)3F&A/CL3I'T'

'/+A '"T*/+&/)"CATI&A&)+/'/A+C*ALLIA&C/

"&5(A "nited &ations (opulation 5und

"&IC/5 "nited &ationsChildrenQs 5und

8* 8orld *ealth rgani4ation


16/129

GI

List of abbreviations

Adm Admission

A&C Antenatal care

( lood pressure

)g )iagnosis

)* )epartment of *ealth

/mC /mergency obstetric care

/&&)+ /arly neonatal death rate

5C 5etal condition

5*+ 5etal heart rate

5'+ 5resh still birth rate

*I *uman immuno%deficiency virus

*rly *ourly

0C 0aternal condition

0CH 0ultiple choice 9uestions

00+ 0aternal mortality ratio

07 0anagement

'C/ b6ective structured clinical e7amination

(CI (erinatal care inde7

(I* (regnancy induced hypertension

(&0+ (erinatal mortality rate

((* (ost%partum haemorrhage

+esus +esuscitation

T Temperature


17/129

1. Background

1.1 ntroduction

/very year women and babies die during childbirth and DE of these deaths occur in the

developing world -8* ;>, p%2. 3lobal efforts place maternal and perinatal mortality at

the centre of their activities and this is evidenced in the 0illennium )evelopment 3oals

-0)3%= and 0)3%>2 that call for a two%third reduction in under%five mortality and a three%

9uarter reduction in maternal mortality ratio -00+2 by the year ;:> using the year ; as a

baseline -Lawn et al ;?, p:==@ (attinson ;?, p:2.

The estimated perinatal mortality rate in ; for the world is = per : births -8* ;?,p:D2, and the estimated maternal mortality ratio in the same year is = per : live births

-Abou$ahr and 8ardlaw ;=, p:;2. The highest numbers of perinatal deaths occur in the

developing world -8* ;?, p;2, and the highest numbers of maternal deaths are almost

e9ually shared between Africa and Asia -Abou$ahr and 8ardlaw ;=, p:@ "&5(A ;=a,

p:;2.

8hile perinatal and maternal mortality estimates are still high mainly in Africa and Asia, 'outh

Africa has made considerable progress in monitoring these deaths with the introduction of the

Confidential /n9uiry into 0aternal )eaths and the implementation of the (erinatal (roblem

Identification (rogramme, resulting in the publication of the 'aving 0others and 'aving

abies reports -(attinson ;?, piv2. These reports reflect on the causes of perinatal and

maternal deaths and provide appropriate recommendations on how to avoid them-(attinson

;?, piv2. )espite national efforts towards alleviating maternal and child deaths, perinatal

mortality in 'outh Africa is still high across the country -(attinson ;?, piv2. The 'outh

African perinatal mortality rate estimate in ; is


18/129

Ta#le /& / MM% estimates per /00 000 li(e #irths #) %egin in 2000&

Regions Maternal mortality ratioper 100 000 live births

Perinatal mortality per1000 births

Global estimates 400 47Africa 830 62Sub-Saharan Africa 920 -South Africa 110 33

Ta#le /& 2 KwaZulu-Natal and Suth A!rica estimated "NM% 2006-2007 per /000

#irths&

Year Perinatal mortality

Ka!ulu-"atal South Africa

#00$ 50.7 38.4

#00% 41.3 38.2

#00& 40.3 34.9

The estimates in Table :.: show high maternal mortality ratios in Africa as a whole but

particularly in sub%'aharan Africa. The table also shows a high maternal mortality ratio for

'outh Africa. Table :.; shows slight decreases in perinatal mortality over a three%year period,

demonstrating a slow move towards the 0)3%= target of a two%third reduction in perinatal

mortality. The decreases shown in perinatal mortality may suggest that the 9uality of care

provided by maternity units in the country is improving. Achieving the 0)3%= target in the

remaining five years is challenging but still possible.

In order to monitor 0)3% = and% >, three indicators are defined for measuring progress towards

the reduction of perinatalPmaternal mortality. The first is the under five%mortality rate -"nited

&ations ;D, p;2, which refers to the probability of dying between birth and e7actly five

years of age. The second is the maternal mortality ratio, referring to the number of maternal

deaths for every : live births -Lawn et al ;?, p:==@ "&5(A ;=a, p:2.The proportion of deliveries attended by skilled health

personnel, therefore, is a key indicator for 0)3% = and% >.

;


19/129

+eductions in perinatal mortality are also dependent on the availability of skilled attendance.

(erinatal deaths may be easier to monitor than maternal deaths, which are relatively infre9uent

events -Akalin et al :, p


20/129

hospital births were attended by un9ualified midwifery assistants without any form of

supervision -#oblinsky et al ;?, p:, which call for a reduction of the (erinatal

and 0aternal 0ortality. 8orldwide, skilled attendance is recogni4ed as the critical factor

towards achieving 0)3s =%> and it is used as one of the indicators to measure the attainment

of these goals. Currently, evidence e7ists in favour of skilled attendance, confirming the

relationship between having a skilled health worker at delivery and the reduction of maternal

and perinatal mortality.

Although the percentage of women attended by health professionals may be increasing

worldwide and in 'outh Africa in particular, (&0+ and 00+ are not being reduced in 'outh

Africa. 0any of these deaths could have been avoided, as most of them were health provider

and administrative related -(attinson et al ;, p;>%;?2. )espite the publication of the 'aving

0others and 'aving abies reports, it has been noticed that the avoidable factors reported in

the first report are the same as reported in recent editions -(attinson ;?, pv2. This is an

indication of no progress and a lack of the implementation of the recommendations provided in

the reports.

0oreover, there is very little information about the competence of skilled attendants -their

knowledge and skills2 as well as the elements that contribute to the enabling environment

within which they work. The Rproportion of deliveries with skilled attendance1 is used as a

pro7y measure for skilled attendance indicating only the presence of a health professional at

delivery -*ussein et al ;=, p:?:2. Therefore, in order to e7plore the issue of skilled

attendance, this study is designed to establish the level of skilled attendance in uThungulu

*ealth )istrict, one district in rural #wa$ulu%&atal.

=


21/129

1.& 'cope of t%e study

This study is an e7ploratory pilot study, e7ploring the relationship between perinatal outcomes,

the 9uality of care provided, the obstetric knowledge and skills of midwives and the enabling

environment. The study comprises five level one hospitals in the uThungulu *ealth )istrict.

This study setting is selected because there is an e7isting 9uality improvement initiative in the

district as part of the Area Three Learning Comple7 initiated by the Centre for +ural *ealth

-C+* ;2. +esults from this study will add to the interventions re9uired to improve learning

activities for health care workers in this district.

0idwives were selected as a focus in the study because they are mostly responsible for the

conduct of normal deliveries and refer in case of complications, especially in rural settings-*arvey et al ;, pD


22/129

1.+'tructure of t%e dissertation

Chapter : has presented the background of the study, the statement of the problem, relevance

and scope of the study, and the aim and ob6ectives of the study. Chapter ; reviews e7isting

literature on skilled attendance, describes different methods used to measure skilled attendance,

and presents the conceptual framework used in the study. Chapter < outlines the methodology

adopted to assess the provision of skilled attendance, the study instruments, the ethical

considerations and study limitations. Chapter = presents the findings of the research in various

formats! statements, tables and graphs. Chapter > discusses the main findings of the study and

makes concluding remarks and recommendations to improve skilled attendance at birth. The

appendices present letters of permissions and the study instruments used for data collection

-Appendices I through III2.

?


23/129

!.Literature review

!.1 ntroduction

This chapter draws together the information obtained from the literature that applies to skilled

attendance at birth and its measurement. It starts with defining skilled attendance and provides

the rationale for skilled attendance at birth. This is followed by the description of different

methods identified in the literature for the assessment of skilled attendance, and the challenges

in, and interventions for, ensuring skilled attendance. The chapter closes with conceptual

frameworks for skilled attendance identified in the literature and the conceptual framework

used in the study.

+elevant literature was identified as follows!

/lectronic databases -0edLine%(ub0ed, 'cience )irect and 3oogle 'cholar2 and

electronic 6ournals -the International Journal of 3ynaecology and bstetrics, +eproductive

*ealth Journal, 'outh African 0edical Journal, 0idwifery@ Tropical )octor and The

Lancet.com2 were searched using the following key words! 'killed attendance at birth A&)

maternal health@ skilled attendance at delivery A&) 9uality maternal care@ skilled

attendance A&) assessment@ skilled attendance A&) measurement@ competency

assessment A&) health care A&) skilled attendance A&) perinatal mortality@ (erinatal

care A&) skilled care@ perinatal mortality A&) health workers@ midwives A&)

knowledge@ midwives A&) skill@ skilled attendance A&) clinical competence.

The 8orld *ealth rganisation website was searched.

/7perts in the field of maternal health were consulted and additional information was

obtained from them.

The reference list at the end of identified literature was used to identify further literature.

Articles provided during the 0aster of (ublic *ealth 0aternal and &ewborn *ealth module

were also used.

The literature was limited to /nglish documents.

The referencing system used is the *arvard referencing style.


24/129

!.! ,%at is skilled attendance-

'killed attendance comprises the presence of a skilled attendant and the enabling environment

in which skilled attendance may be provided -0ac)onagh ;>, p=@ 3raham et al ;:, p:2.

!.!.1 'killed attendants

R'killed attendants1 refers to health professionals providing care to women during childbirth

-8*PIC0P5I3 ;=, p2. These professionals need to have a set of skills, defined as core

midwifery skills, to provide effective care during childbirth. 5or e7ample, they need to be able

to conduct normal deliveries and recogni4e, manage and refer obstetric complications -8*P

IC0P5I3 ;=, p2.

)epending on the level of care -level : or ;2 or geographic location -urban versus rural2 the

skills re9uired might vary in order to respond to the needs of a particular population. *owever,

universally, all skilled attendants are e7pected to perform the core midwifery functions defined

by the "nited &ations -Carlough and 0cCall ;>, p;:@ 3raham et al ;:, p:;%:=2

:. 'afely conduct a normal delivery using aseptic techni9ue

;. Implement active management of the third stage of labour

. 0anually remove the placenta

?. 0anage eclampsia through the provision of parenteral antihypertensives

. +ecogni4e and manage postpartum infection through the use of parenteral antibiotics

D. (erform assisted vaginal delivery through the use of a vacuum e7tractor

. 0anage incomplete abortions with manual vacuum aspiration -0A2

:. #now how to refer women to the ne7t level of care and stabili4e them for their 6ourney.

The types of skilled attendants identified in the literature -8*PIC0P5I3 ;=, p2

comprise!

O 0idwife! a person who has completed the prescribed courses in midwifery and has ac9uired

the license to practice midwifery@O &urse with midwifery skills! a nurse who has ac9uired knowledge and skills in midwifery@

D


25/129

O )octor! a doctor who has ac9uired midwifery skills through training@

O bstetrician! a medical doctor who has speciali4ed in the medical management and care of

pregnancy and childbirth.

!.!.! nabling environment

The enabling environment refers to conditions in which skilled attendants work to provide

women with care during childbirth. The elements of the enabling environment identified in the

literature include the availability of sufficient health professionals, essential e9uipment,

essential drugs, supervision, referral systemsPtransport and a manual of protocols for obstetric

management -3raham et al ;:, p@ 8*PIC0P5I3 ;=, p:=2.

In order to be provided with effective care during childbirth, women need the assistance of

health professionals in numbers proportional to the deliveries in each facility. 5or e7ample, the

+oyal College of bstetricians and 3ynaecologists -+C32 recommend :.:> midwives per

woman in labour -+C3 :, p


26/129

intravenous infusion pump, vacuum e7tractor and suction, obstetric forceps, delivery repair

pack, cervical removal pack, cusco vaginal speculum, symphysiotomy knife, fully e9uipped

resuscitation trolley, defibrillator. The number re9uired of each item depends on the level of

care, the si4e of the labour ward as well as the number of patients treated -artlett et al ;?,

p>D2.

"&IC/5P8*P"&5(A -:, p;?2 propose that the availability of drugs be checked by

whether the signal drugs of in6ectable antibiotics, in6ectable o7ytocics and in6ectable

anticonvulsants are available. 7ytocics are mainly used to reduce the risk of maternal

postpartum haemorrhage, whereas in6ectable antibiotics are used to control mild infections, and

in6ectable anticonvulsants to manage patients with severe pre%eclampsia and eclampsia.

bstetric haemorrhage is the main cause of maternal mortality in the world -8* ;>, p?, p;@ 8* ::, pD@ (admanabhan et al ;, p:;>@ 8*

;D, p:2. In sub%'aharan Africa, the lack of blood for transfusion is shown to be a

contributing factor to maternal mortality. The reason for this deficiency is poverty and a lack of

donors -ates et al ;D, p:2. /nsuring compliance to a manual

:


27/129

of protocols for obstetric management is therefore imperative. This can be achieved through

supportive supervision. It has been shown that regular supervision motivates staff and helps

maintain high standards of care -"&5(A ;?, p:;%:


28/129

The underlying causes of perinatal mortality are, in most cases, the same factors as those

leading to maternal mortality, and include poor health during pregnancy and inade9uate intra%

and postpartum care -0aclean "ndated, p, p;:, )e ernis et al ;, p:, >2. Through effective maternal and fetal monitoring,

skilled attendants can prevent birth asphy7ia. In the case of birth asphy7ia they can

resuscitatePventilate the baby or apply cardiac massage where needed. They can also diagnose

preterm birth and refer for proper care -0aclean "ndated, p


29/129

0oreover, skilled birth attendance at birth is being used as an indicator to measure the

attainment of the 0illennium )evelopment 3oals -*ussein et al ;=, p:?:@ *arvey et al ;,

pD, p2.

The assessment of skilled attendance at delivery has been slow and difficult in many countries

-#oblinsky et al ;?, p:, p;?%;2. Apart from terminology issues, there is also a lack of consistency in the

definition of skilled attendance. 5or e7ample, in 0alawi, ward attendants are considered as part

of skilled attendants, and in &epal, traditional birth attendants were earlier included in the

estimates of the proportion of births by a skilled attendant, but e7cluded later on -3raham et al

;:, p:@ *ussein K Clapham ;>, p;?%;2.

Therefore, in order to distinguish the elements of skilled attendance, various methods are being

used. These include! measuring the presence of skilled attendants -doctors, midwives, and

nurses2, measuring the knowledge and skills of attendants, and measuring the enabling

environment.

!.&.1 /easuring t%e presence of skilled attendants

To be able to achieve two%third and three%9uarter reductions in perinatal and maternal

mortality, enough skilled professionals conducting deliveries are re9uired, and they need to be

accessible. A crude measure of the presence of skilled attendants is the proportion of deliveries

attended by a skilled attendant. To determine the presence of skilled attendants, household

surveys are used. In these surveys women are asked what type of health professional assisted in

their most recent delivery. +esults are presented as the percentage of deliveries by category of

skilled attendant! doctor, midwife, nurse and other. These surveys do not give any indication of

the knowledge or skills of attendants -*arvey et al ;, pD


30/129

A more sophisticated measure has been proposed for the measurement of skilled attendants at

birth! the partnership ratio -3raham et al ;:, p:;=2. This refers to the ratio of deliveries

attended by a doctor versus a midwife. 5rom analysing the association between partnership

ratios and maternal mortality, the ideal partnership ratio -(+2 is :>, D>! where :>E of

deliveries are attended by doctors and D>E of deliveries are attended by midwives -3raham et

al ;:, p:;=@ "&5(A ;E of all deliveries

will result in a complication, and will re9uire higher%level care -0aine et al :, p;=2.

!.&.! /easuring t%e knowledge and skills of attendants

In order to maintain and improve high standards of care, a health system needs regular

assessment of the performance of health professionals. This will help to identify gaps in the

knowledge and skills of providers and the need for training. 0ethods to measure the

performance of health providers include! written tests, computerised tests, review of medical

records and simulations, as reflected in Table ;.< -#ak et al ;:, p:%::2.

Ta#le 2& 6 Methds t assess the per!rmance ! health pr!essinals attending

#irths&

Metho* (hat isassesse* oa*ministere* A*vantages isa*vantages

(ritten test Abilities$traits

andnoledge.

(ase stdies. O tandardisationof )estions.

O o cost.

O !bectivit# in

scoring.

O ills cannot bemeasred.

'omputerise*test

(linicaldecisionmaing sills.

(ase stdies. O (onsistenc# of

the cases.

O !bectivit# in

scoring.

O nabilit# to

evalatecom%etenciesinvolving%h#sical actions

O igh costRevie ofme*icalrecor*s

(om%etence. ecord adit. O roviders arenot aare.

O o cost.

O ncom%letenessof records.

O +issing records

O oor )alit# of

records.

Anatomic

mo*els

(om%etenc#

in %h#sicalactions.

tations. O tandardised

testing.O nabilit# to

simlate.

:=


31/129

!.&.# /easuring t%e enabling environment

The measurement of the enabling environment must include the measurement of the factors

that are essential elements for health worker performance. The assessment can be done by

using methods described in Table ;.= -#ak et al ;:, p::@ 0cCaw%inns et al ;=, p::, :;2.

Ta#le 2& Methds used t measure the ena#ling en(irnment&

Metho* (hat is assesse* o

a*ministere*

A*vantages isa*vantages

Revie ofme*icalrecor*s

erformance ofhealth %roviders inmanaging normaldeliver# and

obstetriccom%lications.

ecord adit. O roviders arenot aare

O o cost.

O ncom%letenessof records

O +issing records

O oor )alit# of

records.

Survey of

health careor)ers

o assess the level

of s%ervision$training$ and

motivation.

,ree listing of

elementscontribting

o the%erformance

of health careorers.

O Assess man#

items at thesame time.

O ime consming

O (om%le" and

difficlt toanal#se.

nventoryof healthcare

facilities

Availabilit# ofessential drgs$e)i%ment and

s%%lies in eachhos%ital.

trctred(heclistsobservation.

O o cost. O :ifficlt to have a

one-si;e fits allcheclist.

!.( 'tudies on t%e measurement of skilled attendance

Although the percentages of deliveries attended by skilled attendants have increased

worldwide, not every health professional can be considered a skilled attendant -*ussein et al

;=, p:?:2. 'ome studies measuring skilled attendance were identified in the literature@

however it seems that most studies measure one or other dimension of attendance. 5ew, with

the e7ception of *ussein et al -;=2, attempt to measure the full dimensions of attendance.

)etails on identified studies measuring the dimensions of skilled attendance are presented in

Table ;.>.

:>


32/129

Ta#le 2& 7Studies n the measurement ! skilled attendance&

Stu*y 2argetgroup

Metho*s use* escription of thestu*y

Results

Mc'a-

.inns et al#00%3

illed birth

attendants.

O s selfassessmentand crrent

sills andnoledge.

O nade)atemonitoring of

labor inmost cases

O nfre)ent

%erformanceof someneborn%ost%artmcare e#tass.

'arloughan* Mc'all

#00&

+aternal andchild health

orers/+(


33/129

!.+ C%allenges in ensuring skilled attendance

This section presents some of the challenges encountered in ensuring skilled attendance.

3lobally, the provision of skilled attendance is slow. There are still gaps in coverage between

developing and developed countries, rich and poor and those living in urban and rural areas

-#oblinsky et al ;?, p:


34/129

!.0 nterventions to increase skilled attendance

To address the factors hindering the provision of skilled attendance, different strategies have

been adopted around the world.

The 8* has passed resolutions re9uesting member countries to develop plans that will

promote retention of health care workers. 5or e7ample, the #ampala )eclaration urges member

countries to ensure ade9uate incentives and a safe working place for health care workers. It also

recommends an e9uitable distribution of health workers across urban versus rural areas -8*%

/C) ;D, p;%;:@ 0ac)onagh ;>, p;, ;:2.

"pgrading health workers1 skills is essential to ensuring women and babies have access to

skilled attendance. This can be achieved by! training nurses in midwifery skills@ training of

general practitioners in obstetric surgery@ or training nurses and midwives in anaesthetic skills

-)e ernis et al ;


35/129

!. Conceptual frameworks for skilled attendance

A number of authors have presented conceptual frameworks to describe the relationship

between different elements of skilled attendance. This section presents the models identified in

the literature and the subse9uent section presents the conceptual framework used in the current

study.

Ta#le 2& 9 5nceptual !ramewrks !r skilled attendance at deli(er)&

'onceptual

frameor)s

escription 5ocus

(,67"5PA67"'856(orl* .an)9 1:::

(onsiders the %artnershi% beteendifferent health %rofessionals able tocare for normal and com%licateddeliveries$ and the enablingenvironment.

ealth %rofessionalsand the enablingenvironment.

(,9 1::% laces silled attendants in theconte"t of a health care centre

s%%orted b# the district hos%ital asthe referral hos%ital in case of

emergenc#.

illed attendants andlevel of care.

Safe Motherhoo* nterAgency Group;Koblins)y #000uipment-Supervision

-Referral6transport-(or)loa*


37/129

This chapter has provided an overview of the e7isting literature on skilled attendance at birth in

general. It has presented the rationale for skilled attendance, providing evidence that skilled

attendants operating within an enabling environment lead to ma6or decreases in maternal and

perinatal mortality. It has also presented measures of skilled attendance and difficulties in the

measurement of skilled attendance. Challenges in ensuring skilled attendance and interventions

to increase skilled attendance were also discussed. 5inally conceptual frameworks for skilled

attendance and the conceptual framework used in the study were discussed.

;:


38/129

#. /et%odology

#.1 ntroduction

This chapter describes the methodology used to measure the provision of skilled attendance in

Level : *ospitals in uThungulu *ealth )istrict. The chapter starts by restating the study aim

and ob6ectives, and proceeds to present the study site, the research design, the study period, the

study population and sampling, the variables measured in the study, the data collection

procedures and instruments, the measures taken to ensure study validity, data management and

storage processes, data analysis, and the ethical considerations in this study. 5inally, the study

limitations are discussed.

#.! $esearc% aim and researc% ob)ectives

The aim of the study was to measure the provision of skilled attendance in Level : *ospitals in

uThungulu *ealth )istrict, with the following ob6ectives!

:. /stablish perinatal outcomes for each Level : *ospital in uThungulu *ealth )istrict.

;. /valuate the 9uality of intrapartum care provided in Level : *ospitals in uThungulu *ealth

)istrict.

. /valuate the environment in which births are attended in Level : *ospitals in uThungulu

*ealth )istrict.

?. Compare the knowledge, skills, environment and 9uality of care with perinatal outcomes.

#.# T%e study site

The study was implemented in the uThungulu *ealth )istrict, located in rural #wa$ulu%&atal.

Its main commercial centre is the port town of +ichards ay. It is bordered by the $ululand

*ealth )istrict in the north, the iLembe *ealth )istrict in the south, the u0khanyakude *ealth

)istrict in the east and the u04inyathi *ealth )istrict in the west.

;;


39/129

"Thungulu *ealth )istrict has a population of DD :< and comprises si7 local authority areas

-#$& )o* ;D2. The *ealth )istrict has two Level ; *ospitals, si7 Level : *ospitals, ==

5i7ed Clinics and := 0obile Clinics, which visit ;>? points. The *ealth )istrict also has si7

local authority clinics -#$& )o* ;D2. As other typical rural districts, uThungulu is isolated

geographically, with poor transport and infrastructure and difficult communication -C+*

;2.

The Level : *ospitals in uThungulu *ealth )istrict are -#$& )o* ;D2!

% Catherine ooth *ospital, which is a :%bed hospital, with a catchment population of

appro7imately ; people.

% /kombe *ospital, situated in a deeply rural area in the &kandla 0agisterial )istrict, off the#ranskop and 'ilutshana main road in the midlands of #wa$ulu%&atal. Ithas ? residential

clinics and < mobile clinics with people.

% /showe *ospital, with =? beds serving a largely rural population estimated to be around


40/129

#.& $esearc% design

This study was an observational descriptive study, which was conducted in two phases!

:2 A retrospective phase, which assessed!O (erinatal outcomes, utilising data from the maternityPdelivery register.

O The 9uality of intrapartum care, through an audit of maternity case records.

;2 A cross%sectional phase, which evaluated!

O The obstetric knowledge of health workers attending deliveries in Level : *ospitals,

utilising a multiple%choice 9uestionnaire -0CH2.

O The obstetric skills of health workers, using an ob6ective structured clinical

e7amination -'C/2.

O The enabling environment, using a maternity unit review form.

#.( 'tudy period

The protocol of the study was submitted to the (ostgraduate /ducation Committee in

&ovember ; and approval was granted on the th0arch ;D -Appendi7 :.:2. It was

submitted to the "niversity1s iomedical +esearch /thics Committee -+/C2 -ref /=:PD2

on the :P2 in August ; and permission for the study to

proceed was received in July ;D, subse9uent to obtaining full ethical approval from +/C

-Appendi7 :.


41/129

#.+ 'tudy population and 'ampling

#.+.1 'tudy population45nits of analysis

This study was located within Level : *ospitals in the uThungulu *ealth )istrict. In order to

meet the ob6ectives of the study there were different units of analysis!

:. To establish perinatal outcomes, the unit of analysis consisted of births in Level :

*ospitals.

;. To assess the 9uality of intrapartum care, the unit of analysis consisted of maternity case

records of women who delivered in Level : *ospitals.


42/129

Ta#le 6& 2 Sampling strateg) per stud) ppulatinunit ! anal)sis ;cnt&2. Three instruments developed by

the )ecentralised (rogramme for Advanced 0idwives were used to test the ability of midwives

to! -:2 use the partograph as a decision%making tool in labour and delivery@ -;2 plot informationon a partograph@ and -2 and marked against a model answer. ne point was awarded for each correct

answer.

In order to assess whether the environment was enabling, the oceP(hilpott facility review

Checklist was used -see Appendi7 ?2, which measured the presence or absence of written

protocols for the management of obstetric complications, the availability of key drugs,

;


46/129

e9uipment and supplies, which are referred to as essential elementsM of obstetric care.

0idwives workload was also assessed. The checklists in each hospital were completed by the

researcher, who reviewed the labour ward with the help of the midwife%in%charge.

Ta#le 6& Summar) ! data cllected, data surce and data cllectin tls&

Bariables ata items ata

source

ata collection

tools

B+,?

*BB:(

otal birthsotal ,?

otal +?otal *BB:otal ?irths G2.5g

:eliver#register

:ata e"tractionform

/see A%%endi" 2

core forAdmission assessment

abor gra%habor management.

ent# five items(overing admission$ labor

gra%h$ labormanagement

+aternit#case

records

hil%ott=Foceabor ecord

evie(heclist /see

A%%endi" 3

Enoledge score formidives> abilit# to

1. (ondct normaldeliver#2. +anage obstetriccom%lications3. +anage F in%regnanc#.

30 )estions on sevento%ics /normal labor$ cord

%rola%se$ %rolonged labor$$ F$ and%er%eral se%sis.

+idives +(D anser sheet/see A%%endi" 4.


47/129

Ta#le 6& Summar) ! data cllected, data surce and data cllectin tls ;cnt abilit# to1. (orrectl# inter%ret

information on a laborgra%h.

2. (orrectl# %lotinformation on a laborgra%h.

3.(orrectl# diagnoseand managing .

*"ercise involving midivesin inter%reting information

on a labor gra%h thiscovered$ ris factors$

diagnosis of %rolongedlabor and +".

*"ercise involving midives

in %lotting information on alabor gra%h. his covered,$ molding$ ca%t$cervical dilation$ descentof head$ fetal %osition$terine contractions$

maternal information$drgs$ maternal ?H %lse$

maternal 0$ rine ot%t.

tation here midivesdiagnosed anddescribed the +" of .

!(*

station

!(*station

!(*station

+ar sheet /see

A%%endi" 5.1

+ar sheet /seeA%%endi" 5.2

+ar sheet /seeA%%endi" 5.3.

*nabling environmentscore based on theavailabilit# of

% rotocols of +"

% rotocol for referral

% *ssential drgs

% *ssential e)i%ment

% %ervision

% +idives> orload

ist of items considered asessential s%%lies$e)i%ment andinfrastrctre for evel 1

os%itals.

!bservation Foce=hil%ott,acilit# eviecheclist /seeA%%endi" 6.

#.2 /easures taken to ensure study validity and control forpotential biases

This section deals with the potential biases in this study and how they were controlled for.

/& Selectin #ias

This could have been introduced in the study design phase. The research utili4ed

convenience sampling by selecting!

% All births in the twelve%month period preceding the commencement of the study

-based on the assumption that births occurring in a facility in a twelve%month period

would not be different from births occurring in any other twelve%month period.


48/129

% 0idwives on duty on the day of data collection -it was assumed that their presence on

the day of data collection was random@ this does not represent a complete picture of

obstetric knowledge and skills of midwives in each hospital2.

% The district within which the study was located. The study site was not randomly

selected@ the choice of the district was made in relation to the e7isting 9uality

improvement initiative in the district, part of the Area Three Learning Comple7

initiated by the Centre for +ural *ealth.

'election bias could have also been introduced in the study implementation phase. The

researcher might have omitted records meeting the inclusion criteria or might have

included those not meeting the inclusion criteria while selecting maternity case records

for review.

$w cntrlled !r

% All Level : *ospitals were included in the study. *owever 0bongolwane hospital

chose not to participate.

% All midwives on duty on the day of data collection were included in the study.

% The information obtained from the study -knowledge and skills2 will not be generalised but

will be treated as pilot study with the prospect for further investigations on a larger scale

-including a representative sample of midwives attending births in Level : hospitals and a

representative sample of Level : hospitals in #wa$ulu%&atal for instance2.

2& *n!rmatin #ias

This might have been introduced in the study design phase by selecting!

% 0aternity case records! data are stored differently in hospitals with some being more

organi4ed than others. In some hospitals, the maternity case records were incomplete, in

some they were missing.

% All midwives on duty on the day of data of data collection! not representative of all

midwives in the district, this will not give the real picture of the knowledge and skills of

midwives in each hospital.

$w cntrlled !r

% Incomplete records were not included in the study.

% The researcher herself collected and analysed data, to avoid information bias.


49/129

% 0idwives and hospitals were identified using a code to avoid mi7ing of results or to

avoid attributing results wrongly. -5or e7ample! symbols identified hospitals,

numbers identified midwives on day shift and letters midwives on night shift2.

% The information obtained from the study will not be generali4ed but will be treated as

pilot study with the prospect for further investigations on a larger scale.

6& Management #ias

Introduced in the study analysis phase, when data is manually captured from 9uestionnaires and

tests, errors may occur.

$w cntrlled !r

All data were stored, captured and analysed in a similar way to avoid management bias. )ata

were doubly entered in /7cel and the 'tatistical (ackage for the 'ocial 'ciencesS -'(''2

version :>. to control for capturing errors.

& Measurement #ias

Introduced in the study design phase@ more than one measurement instruments was used,

adapted from previous studies, whose validity and reliability were not reported.

$w cntrlled !r

% The labour record review form was first piloted at /kombe *ospital on the :>thof July to

test its applicability. &o revision was needed on the study instrument as it could easily be

used.

% The multiple%choice 9uestionnaires and 'C/ 9uestions were piloted at 't 0ary1s *ospital

to test their applicability. &o revision was needed for the multiple%choice 9uestionnaire.

#.13 Data management and data storage

)ouble data entry was performed by the researcher, using 0icrosoft /7cel and the 'tatistical

(ackage for the 'ocial 'ciencesS -'(''2 version :>.. A. &o discrepancies between data were

noted.

)ata was stored on a personal computer, using a password to prevent unauthori4ed access.

ack%up copies of the data were saved on a personal flash disc. The paper tools were kept


50/129

safely in a personal case. )ata will be kept in a safe place until the results of the study are

published.

#.11 Data analysis

The data was analysed using '('' version :>.. The descriptive analytical analyses performed

are described below.

#.11.1 "erinatal outcomes

(erinatal outcomes were calculated as described in 'ection 2. To obtain a

percentage score of the 9uality of intrapartum care, the total number of completed items on the

(hilpottPoce checklist were summed and divided by the total number of items on the

maternity case record, and finally multiplied by one hundred. 'cores were also obtained for

sub%sets by dividing the number of completed items in each subset divided by the total number

of items in the sub%set. #ruskal 8allis &on%(arametric Test -e9uivalent of A&A test2 was

used to determine differences in mean scores and median subset scores between hospitals. The

level of significance was set at .> ->E2. 'eventy percent -E2 was used as an acceptable

score. This was adapted from acceptable scores of performance suggested in the (erinatal

/ducation (rogramme manual.

#.11.# :nowledge assessment

In calculating the obstetric knowledge score of midwives, one point was awarded for each

correct answer. To calculate midwives1 individual knowledge scores, points were summed up

and individual scores were divided by the total number of 9uestions -


51/129

sets2 were also calculated by dividing the total points earned for each sub%set, by the number of

9uestions in each sub%set and multiplied by one hundred. #ruskal 8allis &on%(arametric Test

was used to determine differences in knowledge median scores between hospitals. The level of

significance was set at .> ->E2. /ighty percent -DE2 was used as passing mark. This was

adapted from acceptable scores of performance suggested in the (erinatal /ducation

(rogramme manual.

#.11.& 'kills assessment

In calculating the obstetric skill score of midwives, one point was awarded for each correct

answer. To calculate midwives1 individual total skill scores, individual scores were summed

and divided by the total number of 9uestions and multiplied by one hundred, to obtain a

percentage score. As this data was also negatively skewed, the median and inter9uartile ranges

were used as summary statistic measures. 'cores for each of sub%sets were also calculated by

dividing the points earned for each sub%set, by the number of 9uestions in each sub%set and

multiplied by one hundred. #ruskal 8allis &on%(arametric Test was used to determine

differences in skills median scores between hospitals. The level of significance was set at .>

->E2. /ighty percent -DE2 was used as passing mark. This was adapted from acceptable

scores of performance suggested in the (erinatal /ducation (rogramme manual.

#.11.( nabling environment assessment

Checklists in each hospital were completed by the researcher with the help of the midwife in

charge. The items were dichotomous variables Ryes O no1. ne point was awarded to Ryes1

responses and a score was obtained for each hospital by summing up the number of items that

scored Ryes1 divided by the number of items on the checklist multiplied by one hundred, to

obtain a percentage score for the enabling environment. 'eventy five percent ->E2 was used

as the acceptable overall score for the enabling environment. This was adapted from the study

by oce -;>2.


52/129

#.11.+ Association between perinatal outcomes and 9uality ofintrapartum care8 obstetric knowledge and skills8 and t%e enablingenvironment

)ue to the skewed data, 'pearman1s correlation was used to determine the relationship between

the dependant and independent variables.

The dependent variables were parametric numerical variables defined as!

% (erinatal 0ortality +ate -(&0+2! the number of all perinatal deaths were counted and

added and then divided by the total number of births and multiplied by :.

% 5resh 'tillbirth +ate -5'+2! the number of all fresh stillbirths were counted and added

and then divided by the total number of births and multiplied by :.

% /arly &eonatal )eath +ate -/&&)+2! the number of all early neonatal deaths were divided

by the total number of live births and multiplied by :.

% (erinatal Care Inde7! the perinatal mortality rate was divided by the percentage of low birth

weight babies.

The independent variables were non%parametric 9uantitative, numerical variables. They

are listed and defined below!

% Huality of intrapartum care

The 'core for! Admission assessment, Labour graph, Labour management

% *ealth worker knowledge

The knowledge score of midwives for how to deal with! &ormal delivery, bstetric

complications, *I in pregnancy.

% *ealth worker skills

The skill score of midwives to! Correctly interpret information, plotted on a labour graph,

correctly plot information on a labour graph, correctly diagnose and manage ((*

% /nabling environment

The enabling environment score based on the availability of! (rotocols of management,

(rotocol for referral system, /ssential drugs, /ssential e9uipment, Acceptable midwife

workload, 'upervision.


53/129

#.1! t%ical considerations

#.1!.1 t%ical approval and permissions

All the necessary approvals and permissions were obtained -see section 2. *owever

0bongolwane *ospital did not return a signed form, not confirming nor declining their

participation in the study. A letter was sent on the thctober ;D to the 0bongolwane

*ospital 0anager stating that the consent form had not been received, and that this would be

regarded as declining participation in the study unless otherwise communicated by the *ospital

0anager. &o reply was received. In the letter it was stressed that non%participation would not

carry any negative conse9uences for the hospital -see Appendi7 :.>2.

#.1!.! t%ical principles supported by t%e study

To ensure good 9uality research, the following ethical principles were observed in the design of

the study protocol and its implementation!

% "rinciple ! Autnm)all the necessary information pertaining to the research was made

available and discussed with all hospital managers and midwives, and written informed

consent was obtained from each hospital and midwife -'ee Appendi7 :.?2.

% "rincipal ! #ene!icenceno overt harm was inherent in the design of this study. The

research was designed with the intent of determining the 9uality of skilled attendance in

Level : *ospitals. +evealing a poor 9uality of skilled attendance may carry unintended

negative conse9uences for the hospitals and midwives that participated in the study, for

e7ample if hospital, district or provincial managers use the results of the study in a negative

way. An attempt to manage this will be made in the way that feedback on the study is

provided and in providing recommendations arising from the study. The feedback will

include identifying opportunities for supervision and in%service training.

% 5n!identialit)the information gathered from delivery registers, maternity case records

and midwives remained institutional and not personal! no record was kept of the maternity

case records reviewed. 0idwives and hospitals were identified using symbolic, alphabetical

and numerical coding to ensure their anonymity -e.g. -A2 and A: for the first midwife2.


54/129

% B#ligatin t !eed#ack the results of the study will be fed back to each hospital and to the

district on completion of the e7amination process.

#.1# Limitations of t%e study

3iven logistical and budgetary limitations, this study can only be considered as a pilot study.

This e7ploratory study cannot be considered as representative of hospitals and health care

workers in the province of #wa$ulu%&atal.

The study covered five of the si7 Level : *ospitals in the uThungulu *ealth )istrict. It would

have been of greater value if 0bongolwane *ospital had agreed to participate in the study.

Their non%inclusion in the study does not allow a complete picture of skilled attendance in

Level : *ospitals in the uThungulu *ealth )istrict as a whole.

In the audit of maternity case records, there were missing files and incomplete records

particularly at /showe, &kandla and 't 0ary1s *ospitals. This limited the proportion of

records that could actually be reviewed.

)ue to the shortage of midwives in Level : *ospitals, the knowledge and skills tests were

administered in the midst of normal duty. 'ome labour wards were very busy. This may have

affected the 9uality of response by midwives.

Convenience sample of midwives was used in the study rather than a survey. This may have

affected the internal validity of the study. It is not possible to determine to what degree

midwives in this sample are representative of all midwives in each Level : *ospital in the

uThungulu *ealth )istrict.5urthermore, the study could not control for potential confounders

such as! differences in training, years of e7perience between midwives and within hospitals.

#.1& Conclusion

This section described the methodology used in this study. It restated the study aim and

ob6ectives. It presented the research aim and ob6ectives, the study site, the study design, the

study period, the study population and sampling. It went on to describe the variables measured

in the study, the data collection procedures and instruments, the measures taken to ensure study


55/129

validity, data management and storage processes, data analysis, and the ethical considerations

in this study. 5inally, it discussed the study limitations.


56/129

&. $esults

This chapter presents the results of the study according to the study ob6ectives. The main aim of

the study was to measure the provision of skilled attendance in Level : *ospitals in uThungulu

*ealth )istrict, with the following ob6ectives!

:2 /stablish perinatal outcomes for each Level : *ospital in uThungulu *ealth )istrict.

;2 /valuate the 9uality of intrapartum care provided in Level : *ospitals in uThungulu

*ealth )istrict.

2 /valuate the environment in which births are attended in Level : *ospitals in


?2 Compare the 9uality of care, the knowledge, skills and environment with perinatal

outcomes.

The chapter starts by presenting the perinatal outcomes calculated for each hospital, and then

proceeds to present the 9uality of intrapartum care, the results of the knowledge and skills tests,

and the results on the enabling environment and workload of midwives. 5inally, the association

is measured between perinatal outcomes and 9uality of care, obstetric knowledge and skills,

and the enabling environment.

&.1"erinatal outcomes

Table =.: presents raw perinatal data -total number of births, live births, fresh stillbirths,

macerated stillbirths and low birth%weight births2 by hospital, from July ; to June ;D.

=


57/129

Ta#le & /"erinatal data #) @e(el / $spital uThungulu $ealth District, Cul) 200

t Cune 2008&

ospital 2otal

births

2otal live

births

5S. MS. 8"" C.(

'3.ooth 368 358 1 9 7 428)ombe 722 701 11 10 13 978shoe 2520 2472 22 26 17 257")an*la 1592 1572 1 19 26 168St Mary 1335 1309 3 23 13 1332otal 6537 6412 38 87 76 697

5rom these data perinatal outcome indicators were calculated and presented in Table =.;. The

perinatal mortality rate -(&0+2 for the five hospitals combined was


58/129

Ta#le & 6 >?planatr) data n the actual sample si=e ! maternit) case recrds

audited in each @e(el / $spital in uThungulu $ealth District&

ospitals

'3.ooth 8)ombe 8shoe ")an*la St MaryBmber of admissions 32 58 183 168 146ecords for inclsion 18 30 83 130 56Admitted not monitored 0 0 25 2 16

otal records e"clded *"cl (ervical dil.C8cm

*"cl ??A*"cl (=

*"cl +issing files

14 28 75 36 746 12 26 19 142 7 19 8 1660

30

624

90

1232

otal records revieed 18 30 83 112 56

In the following section, the results of the maternity case record review are presented! firstly,

presented is the overall percentage of maternity case records with each item recorded, and

secondly, the score per subset in the maternity case record, by Level : *ospital. The details are

reflected in Tables =.= and =.>.

Ta#le & "ercentage ! maternit) case recrds with each item recrded, #) @e(el /

$spital in uThungulu $ealth District, Cune 2008&

tems '3.oothnD1?

8)ombenD$0

8shoenD?$

")an*lanD11#

StMary

nD&E

2otalnD#::

A*missionassessment

AB( card revieed 22 83 89 74 64 74Adm. form com%lete 61 93 100 100 100 97:iagnosis H +" 61 93 100 99 98 96Adm. doble checed 0 0 0 0 0 0

Cabourgraph

is factors recorded 55 93 98 95 73 90, I hrl# 88 100 100 99 91 97tate of li)or 83 100 100 99 89 97:egree of molding 83 100 100 100 85 96(ontractions I hrl# 100 100 100 100 96 99:ilatation 100 100 100 100 100 100:il.-correct %lotting 100 100 100 100 100 100

evel of head 4 hrl# 94 100 100 100 100 99+aternal ? 94 96 100 99 80 95+aternal 0 94 96 98 94 78 91ecord of drgs H flids 61 70 72 98 50 72

=;


59/129

Ta#le & &"ercentage ! maternit) case recrds with each item recrded, #) @e(el /

$spital in uThungulu $ealth District, Cune 2008 ;cnt


60/129

Table =.> presents maternity case records subset scores by Level : *ospital in uThungulu *ealth

)istrict. The subsets include! the admission assessment, the labour graph and the management of

labour. The labour graph was further divided into the monitoring of fetal condition -5C2,

monitoring of labour progress -L(2 and monitoring of maternal condition -0C2. The table shows

variable median scores between hospitals in the admission assessment -range


61/129

5igure =.; shows the median, ma7imum and minimum values and inter%9uartile range for the

overall knowledge test scores by hospital. It highlights differences in scores between providers

within and amongst hospitals. The lowest individual knowledge score -E2 at Catherine ooth and 't 0ary1s hospitals

respectively. The lowest overall median score -=DE2 is recorded at &kandla *ospital and the

highest overall median score ->E2 at 't 0ary1s *ospital. #ruskal 8allis &on%(arametric Test

shows no statistically significant difference in the overall knowledge scores between hospitals

-pB.2.

:igure & 2Knwledge median percentage scres #) @e(el / $spital in uThungulu

$ealth District, Cune 2008&

Table =.? presents the knowledge median subset scores by hospital, e7pressed as a percentage. It

also presents the difference in knowledge median subset scores by means of #ruskal 8allis &on%

St Mary")an*la8shoe8)ombe'3.ooth

Knole*gem

e*ian

+

scoresperho

spi

tal

70

60

50

40

30

=>


62/129

(arametric Test, which shows no statistically significant differences in the knowledge median

subset scores amongst hospitals.

Ta#le & 9 B(erall knwledge median su#set scres and di!!erence in knwledge median

su#set scres, #) @e(el / $spital in uThungulu $ealth District, Cune 2008&

'3.oothnD$

8)ombenD#

8shoenDE

")an*lanD:

St MarynD&

'hi-s>uare

P-value

"ormal labour 42 43 43 42 52 4.1 0.38Prolapse* cor* 67 50 58 39 50 1.88 0.75Prolonge* labour 100 75 75 72 80 2.87 0.57PP 44 50 33 56 73 4.30 0.36B 78 83 72 74 80 0.44 0.97P 67 33 28 37 60 7.09 0.34Sepsis 33 50 67 44 50 4.44 0.38

&.&*bstetric skills of %ealt% care workers

Twenty%five midwives from the five hospitals completed the test of obstetric skills. The breakdown

of the number of midwives by hospital is presented in Table in Chapter


63/129

&.&.! Labour E2 was

recorded at 't 0ary1s *ospital. #ruskal 8allis &on%(arametric Test shows no statistically

significant differences in the overall Labour 3raph /7ercise I median scores between the five

hospitals -pB.>?2.

Table =.D shows the Labour 3raph /7ercise I median subset scores and differences in the median

subset scores per hospital. 't 0ary scored highest on identification of risk factors -?E2, Catherine

ooth on diagnosis -?E2 and /kombe on management ->E2. There are statistically significant

differences in the median scores for the identification of risk factors -pB.;>2 and diagnosis

-pB.:2.

Ta#le & 8 @a#ur raph >?ercise * su#set median scres and di!!erences in su#set

median scres, #) @e(el / $spital in uThungulu $ealth District, Cune 2008&

'3.oothnD$

8)ombenD#

8shoenDE

")an*lanD:

St MarynD&

'his>uare

p-value

Ris) factors 69 57 61 69 76 11.1 0.025iagnosis 100 50 100 50 50 17.6 0.001M/ 14 50 37 37 42 6.10 0.192

=


64/129

:igure & 6 @a#ur raph >?ercise * median percentage scre #) @e(el / $spital inuThungulu $ealth District, Cune 2008&

&.&.# Labour


65/129

:igure & @a#ur raph >?ercise ** E median percentage scre #) @e(el / $spital in

uThungulu $ealth District, Cune 2008&

The Labour 3raph /7ercise II subset scores are presented in Table =.:. It shows high scores in all

five hospitals in different subsets. It also shows no statistically significant difference in median

subset scores amongst hospitals.

Ta#le & F@a#ur raph >?ercise **- median su#set scres and di!!erences in median

su#set scres, #) @e(el / $spital in uThungulu $ealth District, Cune 2008&

'3.oothnD$

8)ombenD#

8shoenDE

")an*lanD:

St MarynD&

'his>uare

p-value

5etal con*ition 63 67 63 70 73 9.57 0.48Progress of labour 80 80 66 80 80 6.11 0.19Maternal con*ition 75 67 80 100 80 5.40 0.24

CabourGraph8/ercise

-me*ian+sc

oreper

hospital

St Mary")an*la8shoe8)ombe'3.ooth

80

70

60

50

=


66/129

&.&.& "ost> partum %aemorr%age station

5igure =.> shows differences in scores recorded for the ((* station, for midwives within and

amongst hospitals. The lowest individual ((* score -


67/129

:igure & 7 ""$ Statin - B(erall median percentage scre, #) @e(el / $spital in

uThungulu $ealth District, Cune 2008&

St MaryFs")an*la8shoe8)ombe'3.ooth

80

70

50

40

30

60

PP0stationMe*ian+s

corep

erhospital

>:


68/129

Post- partum haemorrhage station

5igure =.> shows differences in scores recorded for the ((* station, for midwives within and

amongst hospitals. The lowest individual ((* score -na#ling en(irnment scres, #) @e(el / $spital in uThungulu $ealth

District, Cune 2008&

'3.ooth 8)ombe 8shoe ")an*la StMary

2otal

Protocols of management 1 1 1 1 1 1008mergency response 0 0 1 0 0 20Pac)e* cells 1 0 0 1 1 405reeHe *rie* plasma 0 0 1 1 1 60nectable antibiotic 1 1 1 1 1 100nectable anticonvulsants 1 0 1 1 1 80nectable o/ytocics 1 1 1 1 1 100

At least # .P machines 1 0 0 0 1 40At least # '2G machines 1 0 1 1 1 80At least # stethoscopes 1 1 1 1 1 1005etoscope 1 1 1 0 1 80optone 0 0 0 1 1 60

At least # Bacuum e/tractor 1 1 1 0 1 80"eonatal resus3 facilities 1 1 1 1 1 100,perating theatre 1 1 1 1 1 100

Supervision on both shifts by m6 1 1 1 0 1 80Supervision on *ay shift by AM 1 1 1 0 1 80Supervision on night shift by AM 0 1 0 0 0 20

;


69/129

Acceptable or)loa* 1 1 0 0 1 40

2otal 15 12 16 11 16 14

+ 78 63 84 57 84 74

Table =.:: presents the enabling environment subset scores by hospital. The subsets include!

referral, drugs and supplies, e9uipment, supervision and workload. The table shows variable scores

between hospitals in the referral subset ->E%:E2, in the scores for drug and supplies -=E%

:E2, supervision and workload -E%:E2. *owever, the table shows that the range is narrower

for the e9uipment subset -?;E%:E2.

>


70/129

Ta#le & //>na#ling en(irnment su#set scres, #) @e(el / $spital in uThungulu

$ealth District, Cune 2008&

'3.ooth 8)ombe 8shoe ")an*la St Mary

Referral 50 50 100 50 50rugs an* supplies

100 40 80 100 1008>uipment 87 62 75 62 100Supervision 33 100 33 0 33(or)loa* 100 100 0 0 1

Table =.:; presents midwives1 workloadby hospital for June ;D. Three out of five hospitals

have an acceptable workload ratio -:.:> midwives per woman in the labour ward2 as recommended

by the +C3 -:, p


71/129

enabling environment. Table =.:< shows the 'pearman1s correlation coefficients and p%values. All

were neither clinically nor statistically significant.

Ta#le & /6 Spearman1s crrelatin #etween dependent and independent (aria#les&

&.0Conclusion

This chapter has presented the findings of the study, analysed using descriptive and inferential

statistics. verall and subset scores were calculated where appropriate. )escriptive statistics were

used to provide indicators of perinatal care, a summary of the 9uality of intrapartum care

-revealing high mean overall scores in the different hospitals2, the obstetric knowledge and skills

of midwives -showing variable median scores between hospitals2 as well as the enabling

environment and the workload of midwives.

#ruskal%8allis Tests was used to determine differences in median labour record subset between

hospitals showing statistically significant differences of scores amongst hospitals. #ruskal%8allis

Test was used to determine differences in knowledge and skill median scores amongst hospitals,

Iuality ofintrapartum

care

,bstetric)nole*ge

,bstetric s)ills 8nablingenvironment

rho p rho p rho p rho pP"MR -0.35 0.55 0.10 0.87 0.00 1.00 -0.20 0.745S.R 0.50 0.39 0.10 0.87 0.60 0.28 -0.10 0.84

8""R -0.60 0.28 0.60 0.28 -0.50 0.39 -0.20 0.74

>>


72/129

showing no statistically significant difference in the overall knowledge scores amongst hospitals.

*owever, a statistically significant difference was detected in the median overall skill score

amongst hospitals.

'pearman correlations were used to determine relationships between perinatal outcomes, the

9uality of intrapartum care and obstetric knowledge and skills of midwives and the enabling

environment and no correlation between variables was found.

>?


73/129

(.Discussion

The main aim of the study was to measure the provision of skilled attendance in Level : *ospitals

in the uThungulu *ealth )istrict. (erinatal outcomes were assessed as indicators of the

effectiveness of skilled attendance and as a measure of the 9uality of intrapartum care. The

dimensions of skilled attendance that are discussed include! the 9uality of intrapartum care, the

knowledge and skills of midwives, and the enabling environment. The overall provision of skilled

attendance is also discussed. Implications for practice, interventions in the health system and for

further research are identified.

(.1 "erinatal outcomes

There is controversy as to whether perinatal mortality rates can be used as pro7y measures for

maternal mortality. 0aternal deaths are infre9uent and cannot therefore be used effectively for

measuring the impact of skilled attendance. A high proportion of perinatal mortality can be averted

by the provision of skilled attendance at birth. Therefore perinatal care indicators remain effective

measures of the 9uality of care, particularly the perinatal care inde7 -(attinson et al ;, p>2.

verall the (&0+ -2 includes an

indicator of 5resh 'till irths U /arly &eonatal )eaths -)ay :2 rate while the previous 'aving

abies ; report -0+C +esearch "nit for 0aternal and Infant *ealth Care 'trategies et al

"ndated, p2 reports on the overall 'till irth +ate. The more comparable inde7 may be the

5'U/&&) -):2 rate, which is reported to be :?.; per : for district hospitals nationally

-(attinson et al ;, p>2. The 5'+ -:> per :2 for /kombe hospital appro7imates this, while

the 5'+ for all four other hospitals is below : per : births, with three -Catherine ooth,

&kandla and 't 0ary2 showing 5'+s of less than > per :.

>


74/129

The combined /&&)+ for the five hospitals is :; per : live births. Compared with the

national -:;.> per : live births2 and provincial rate -:>.= per : live births2 -(attinson et al

;, p>2 most hospitals perform poorly, with Catherine ooth, /kombe and &kandla hospitals all

showing /&&)+' of :%; per : live births.

A more helpful analysis of early neonatal deaths would be separating those that die on )ay : post

delivery versus those that die later in the first week of life. )eaths on )ay : are measure of the

9uality of intrapartum care and deaths from day ;% are a measure of the 9uality of newborn care.

The 'aving abies report ;?%; proposes a new inde7 for measuring the 9uality of

intrapartum care! the 5'U/&&) -:d2 -(attinson et al ;, p=2. It must be assumed that higher

/&&)+s in the uThungulu )istrict *ospitals, in the light of very low 5'+s, represent a high

proportion of early neonatal deaths on day one. This may apply in Catherine ooth, &kandla and

't 0ary1s hospitals. *owever, in /kombe *ospital, both the 5'+ and the /&&)+ are very high,

indicating both poor intrapartum and newborn care. In /showe *ospital, both rates are low,

indicating good intrapartum and newborn care.

The (CI in all hospitals ranged between < and =, and the combined (CI reported for the hospital in

the district is 2. The perinatal care inde7 is a true measure of the

9uality of intrapartum careM -(attinson et al ;, p>2 and poor performance against this inde7

suggests that a high proportion of babies of good weight are dying. These deaths could be averted.

5rom the measure of perinatal outcomes in the uThungulu *ealth )istrict, it appears that three

hospitals -/showe, &kandla and 't 0ary2 are performing generally well. ut the (CI still indicates

a poor 9uality of care and thus all have deaths that could be avoided. 5urther research is

recommended to study the causes and avoidable factors in these deaths. &ational surveys

-(attinson et al ;, p :>2 show that in district hospitals, 6ust over ;E of deaths could be

associated with the health care provider and 6ust under :>E with administrative problems,

indicating problems both with the skilled attendants and the enabling environment.

>D


75/129

(.! ;uality of intrapartum care

0aternity case records are the main source of information in the labour ward. They contain

antenatal, intrapartum and postpartum information of women admitted in labour. 0any times

information on care is recorded in maternity case records but not performed. *owever, there are

cases where care is not documented but was performed -'andin%o6o et al ;?, p2. 8hen

auditing the 9uality of care, if the information is not recorded, it must be assumed that the

procedure was not performed -*ussein et al ;=, p:??2.

All hospitals have high overall mean percentage scores per record. In the study by oce -;>2

scores of E were only reached after an intervention was implemented@ whereas in the study by

*ussein et al -;=2 most items of the maternity record were not recorded, although they did not

measure the same parameters used in the current study. The #ruskal%8allis &on%(arametric Testshowed a statistically significant difference in scores amongst hospitals, suggesting that all five

hospitals do not perform at a similar level in terms of the 9uality of care provided.

(oor scores on the admission assessment suggest that any risk factors present during antenatal care

may not be recogni4ed on admission and the appropriate plan for delivery may not be made. This

calls for further research to investigate the relationship between the 9uality of admission

assessment and perinatal outcomes. This may result in the need for in%service training and the need

for a supervision intervention.

All hospitals scored similarly well on the recording of findings on the labour graph. This is not

consistent with the findings by *ussein et al -;=2! their study reports only :>.=E of completed

labour graphs.

It is noted that in all hospitals, the management of labour scored most poorly. This is consistent

with the findings by 3bangbade et al -;


76/129

result in need for training and supervision interventions to improve the capacity of midwives to

interpret findings and make management decisions.

Comparison of the subsets with #ruskal%8allis &on%(arametric test shows statistically significant

differences in scores amongst hospitals. This suggests that labour monitoring is not performed at

the same level in all five hospitals.

(.# *bstetric :nowledge

verall, all hospitals scored poorly on the knowledge test. &o hospital met the (/( standard

of DE. These results compare with *arvey et al -;2, 3bangbade et al -;


77/129

scores in interpreting information on a labour graph. This is consistent with the findings of the

labour record review where low scores were found on the management of labour.

verall, there was not a statistically significant difference in median scores on the Labour 3raph

/7ercise I and II amongst hospitals, indicating that midwives1 obstetric skills -in plotting and

interpreting information on a labour graph2 in the study hospitals are similar. *owever, there was a

statistically significant difference in the post%partum haemorrhage median scores amongst

hospitals suggesting that hospitals perform on a different level in terms of the management of post%

partum haemorrhage.

Increasing the proportion of deliveries assisted by skilled attendants is the key strategy towards

reductions in maternal and perinatal mortality. There are a number of interventions that can be

performed by skilled attendants to decrease maternal and perinatal mortality and these are listed in

section ;.< -Table ;.:2. *owever, the above results demonstrate that there is a need for reviewing

the education and training received by midwives, as suggested by the literature in section ;.?.

(oor scores on plotting and interpreting information on a labour graph suggest that there are

difficulties in the ability of midwives to transfer knowledge to skills. It also indicates differences in

abilities to record versus interpret findings on the partograph for the management of labour. This

has similar implications as the poor scores in the management of labour! there is a need for training

and supervision interventions to improve the capacity of midwives to interpret findings and make

management decisions.

There are a number of interventions that can be performed by skilled attendants to decrease

maternal and perinatal mortality and these are listed in section ;.< -Table ;.:2. *owever, the above

results demonstrate that there is a need for reviewing the education and training received by

midwives, as suggested by the literature in section ;.?. (oor scores on plotting and interpreting

information on a labour graph suggest that there are difficulties in th

Date post:	13-Apr-2018
Category:	Documents
Upload:	solange-zoe
View:	219 times
Download:	0 times

Measuring Skilled Attendance in the uThungulu District 2 Mar 2

Documents