Prioritising gender, equity, and human rights in a GRADE ... · equity and human rights (GER) in the provision of self- care interventions is key to ensuring better health for all.

Page 1

1Siegfried N, et al. BMJ Global Health 2020;5:e002128. doi:10.1136/bmjgh-2019-002128

Prioritising gender, equity, and human rights in a GRADE- based framework to inform future research on self care for sexual and reproductive health and rights

Nandi Siegfried ,1 Manjulaa Narasimhan,2 Carmen H Logie,3 Rebekah Thomas,4 Laura Ferguson,5 Kevin Moody,6 Michelle Remme7

Original research

To cite: Siegfried N, Narasimhan M, Logie CH, et al. Prioritising gender, equity, and human rights in a GRADE- based framework to inform future research on self care for sexual and reproductive health and rights. BMJ Global Health 2020;5:e002128. doi:10.1136/bmjgh-2019-002128

Handling editor Seye Abimbola

Received 30 October 2019Revised 28 January 2020Accepted 15 February 2020

For numbered affiliations see end of article.

Correspondence toDr Nandi Siegfried; nandi. siegfried@ gmail. com

© Author(s) (or their employer(s)) 2020. Re- use permitted under CC BY. Published by BMJ.

AbsTrACTIntroduction In January 2019, the WHO reviewed evidence to develop global recommendations on self- care interventions for sexual and reproductive health and rights (SRHR). Identification of research gaps is part of the WHO guidelines development process, but reliable methods to do so are currently lacking with gender, equity and human rights (GER) infrequently prioritised.Methods We expanded a prior framework based on Grading of Evidence, Assessment, Development and Evaluation (GRADE) to include GER. The revised framework is applied systematically during the formulation of research questions and comprises: (1) assessment of the GRADE strength and quality rating of recommendations; (2) mandatory inclusion of research questions identified from a global stakeholder survey; and (3) selection of the GER standards and principles most relevant to the question through discussion and consensus. For each question, we articulated: (1) the most appropriate and robust study design; (2) an alternative pragmatic design if the ideal design was not feasible; and (3) the methodological challenges facing researchers through identifying potential biases.results We identified 39 research questions, 7 overarching research approaches and 13 discrete feasible study designs. Availability and accessibility were most frequently identified as the GER standards and principles to consider when planning studies, followed by privacy and confidentiality. Selection and detection bias were the primary methodological challenges across mixed methods, quantitative and qualitative studies. A lack of generalisability potentially limits the use of study results with non- participation in research potentially highest in more vulnerable populations.Conclusion A framework based on GRADE that includes stakeholders’ values and identification of core GER standards and principles provides a practical, systematic approach to identifying research questions from a WHO guideline. Clear guidance for future studies will contribute to an anticipated ‘living guidelines’ approach within WHO. Foregrounding GER as a separate component of the framework is innovative but further elaboration to operationalise appropriate indicators for SRHR self- care interventions is required.

InTroduCTIonThe mandate of the WHO is to develop global clinical and public health guidance that informs country- level healthcare policies, guidelines, programmes and services. The systematic process of WHO guideline devel-opment includes the identification and docu-mentation of research gaps.1 This is intended to focus research to inform and strengthen

Key questions

What is already known? ► Reliable methods to identify research gaps from WHO guidelines are lacking with gender, equity and human rights (GER) infrequently prioritised when for-mulating future research questions.

What are the new findings? ► A prior Grading of Evidence, Assessment, Development and Evaluation (GRADE)- based frame-work was revised to include mandatory stakeholder input and selection of relevant GER standards and principles and successfully applied to a WHO guide-line on self- care interventions for sexual and repro-ductive health and rights (SRHR).

► 39 research questions, 6 research approaches and 16 discrete study designs for SRHR self- care were identified and the key methodological issues outlined.

► Availability and accessibility were the GER standards and principles most frequently identified as relevant.

What do the new findings imply? ► Application of the GRADE- informed framework to future WHO guidelines development processes has potential to harmonise research question formula-tion and to ensure more consistent consideration of GER across the organisation.

► The dashboard outlining future SRHR studies is ex-pected to prove useful to researchers.

on July 28, 2020 by guest. Protected by copyright.

http://gh.bmj.com

/B

MJ G

lob Health: first published as 10.1136/bm

jgh-2019-002128 on 30 March 2020. D

ownloaded from

Page 2

2 Siegfried N, et al. BMJ Global Health 2020;5:e002128. doi:10.1136/bmjgh-2019-002128

BMJ Global Health

new and existing WHO recommendations. However, wide variation in the quality, robustness, presentation and dissemination of research priorities across sets of WHO guidelines has been noted.2 Currently within the WHO transformation process, systematic approaches to coherent research agenda formulation are being explored.

In 2018, the WHO Department of Reproductive Health and Research embarked on a process to: (1) develop evidence- based recommendations on self- care inter-ventions for sexual and reproductive health and rights (SRHR) and (2) delineate research gaps in the field. A Guidelines Development Group (GDG) was established comprising experts and stakeholders active in SRHR and/or self- care interventions. The GDG was regionally and gender representative and included representation from youth, vulnerable populations, healthcare providers, policymakers, programme managers, researchers and civil society, as well as experts in human rights, gender equality and health economics.3

The following definition of self- care was adopted for the guidelines: self- care is the ability of individuals, families and communities to promote health, prevent disease, maintain health and to cope with illness and disability with or without the support of a healthcare provider.4 The scope of self- care as described in this definition includes health promo-tion, disease prevention and control, self- medication, providing care to dependent persons, seeking hospital/specialist care if necessary and rehabilitation including palliative care.

Members of the GDG attended a scoping meeting in March 2018 to prioritise and refine the self- care inter-ventions to be evaluated in the guideline and to reflect on discussions arising from a prior WHO- led meeting on the ethical, legal, social accountability and human rights implications of self- care interventions.5 The GDG agreed that the systematic consideration of gender, equity and human rights (GER) in the provision of self- care interventions is key to ensuring better health for all. As defined by WHO, human rights- based approaches to health consider and address inequalities, power imbal-ances and discrimination, including those related to gender.6 Such approaches also aim to support better and more sustainable health and development outcomes and focus on capacity development, both of duty bearers to meet their obligations and of individuals to claim their rights.7 Attention to ways of empowering individuals and communities, particularly vulnerable populations, to understand and claim their rights requires special atten-tion in many domains, including in research. Sridharan and colleagues8 argue that WHO needs to develop concrete actions towards mainstreaming GER. The GDG selected the fulfilment of the following GER standards and principles as key to measuring the impact of self- care interventions in future research studies: (1) the right to highest attainable standard of health (including avail-ability, accessibility, acceptability and quality); (2) active and fully informed participation; (3) non- discrimination;

(4) the right to seek, receive and impart information; (5) informed decision- making; (6) privacy and confidenti-ality; and (7) accountability.9

At a final GDG meeting in January 2019, the GDG reviewed evidence from five systematic reviews on self- care interventions in order to formulate new consensus- based self- care recommendations.3 The new recommendations covered the following topics: (1) self- injectable hormonal contraception, (2) over- the- counter oral contraception, (3) home- based ovulation predictor kits (OPKs), (4) self- sampling for human papilloma virus (HPV) and (5) self- collection of samples for sexually transmitted infections (STIs). Each recom-mendation was formulated in response to an a priori clinical or public health question regarding the effec-tiveness of the intervention when offered as an addi-tional approach to current practice within the formal health sector. The questions were formulated using the Population, Interventions, Comparison and Outcomes (PICO) structure.1

To further inform the guidelines, the GDG also reviewed evidence from a global online survey of health-care providers and users of healthcare services, which was hosted on the WHO website and shared via several listservs between July and October 2018.10 The survey included a range of questions regarding respondents’ values and preferences for self- care interventions for sexual and reproductive health. Three of these were interventions evaluated in the systematic reviews: (1) self- injectable hormonal contraception, (2) oral contracep-tion (including over the counter), and (3) self- collection of samples for STIs (including HPV). The survey reached 294 (35.6%) healthcare providers and 531 (64.4%) users of healthcare services from 113 countries. There was diversity in WHO regional representation in responses from healthcare providers (Africa 30.0%, Europe 20.8%, Latin America and the Caribbean 20.8%, Asia 14.2%, Northern America 13.9% and Oceania 0.3%) and lay respondents (Europe 32.0%, Africa 23.0%, Asia 20.2%, Northern America 12.9%, Latin America and the Carib-bean 10.5% and Oceania 1.3%).10

In this article, we present a systematic approach to the formulation of research questions to guide future studies related to the five topics of the new WHO recommen-dations on self- care interventions. We demonstrate how evidence from systematic reviews, results from a global survey and consideration of GER, together informed question formulation. Our primary aim is to demon-strate both the feasibility and utility of using a structured process combined with a GER lens to identify research gaps within the context of developing a WHO guideline. A secondary aim is to provide researchers with a dash-board of potential evidence- informed research ques-tions and related feasible study designs—and associated methodological challenges and GER considerations—to ensure responsive future research in the rapidly evolving field of SRHR self- care.

on July 28, 2020 by guest. Protected by copyright.

http://gh.bmj.com

/B

MJ G

lob Health: first published as 10.1136/bm

jgh-2019-002128 on 30 March 2020. D

ownloaded from

Page 3

Siegfried N, et al. BMJ Global Health 2020;5:e002128. doi:10.1136/bmjgh-2019-002128 3

BMJ Global Health

Table 1 Key domains that require consideration when formulating WHO recommendations

Factor How the factor influences the direction and strength of a recommendation

Quality of the evidence The quality of the evidence across outcomes critical to decision making will inform the strength of the recommendation. The higher the quality of the evidence, the greater the likelihood of a strong recommendation.

Values and preferences This describes the relative importance assigned to health outcomes by those affected by them; how such importance varies within and across populations; and whether this importance or variability is surrounded by uncertainty. The less uncertainty or variability there is about the values and preferences of people experiencing the critical or important outcomes, the greater the likelihood of a strong recommendation.

Balance of benefits versus harms

This requires an evaluation of the absolute effects of both benefits and harms (or downsides) of the intervention and their importance. The greater the net benefit or net harm associated with an intervention or exposure, the greater the likelihood of a strong recommendation in favour or against the intervention.

Resource implications This pertains to how resource intense an intervention is, whether it is cost–effective and whether it offers any incremental benefit. The more advantageous or clearly disadvantageous the resource implications are, the greater the likelihood of a strong recommendation either for or against the intervention.

Priority The problem’s priority is determined by its importance and frequency (ie, burden of disease, disease prevalence or baseline risk). The greater the importance of the problem, the greater the likelihood of a strong recommendation.

Equity and human rights The greater the likelihood that the intervention will reduce inequities, improve equity or contribute to the realisation of one or several human rights as defined under the international legal framework, the greater the likelihood of a strong recommendation.

Acceptability The greater the acceptability of an option to all or most stakeholders, the greater the likelihood of a strong recommendation.

Feasibility The greater the feasibility of an option from the standpoint of all or most stakeholders, the greater the likelihood of a strong recommendation. Feasibility overlaps with values and preferences, resource considerations, existing infrastructures, equity, cultural norms, legal frameworks and many other considerations.

Reproduced from the WHO 2014.1

MeTHodsWe adopted a similar methodological approach to research formulation used in two previously published WHO guidelines.11 12 Prior to the GDG meeting, a system-atic review, including meta- analysis where appropriate, was conducted for each of the five selected PICO ques-tions.13–17 The overall certainty of evidence was rated as high, moderate, low or very low according to the Grading of Evidence, Assessment, Development and Evalua-tion (GRADE) approach.18 The certainty of evidence is dependent on the risk of bias, precision, consistency, directness of the results and other considerations such as publication bias. During the meeting, the GDG formu-lated a recommendation in response to the following GRADE domains: certainty of the evidence, balance of benefits and harms, resource use implications, user values and preferences, acceptability among healthcare providers and key stakeholders, feasibility, equity and human rights (See table 1). Recommendations were then further categorised by the GDG as strong or conditional. In general, strong recommendations are made when the quality of evidence is high and the benefits of an intervention clearly outweigh the harms, whereas condi-tional recommendations recognise that the quality of the

evidence is low or that specific country contextual factors may determine the uptake of a recommendation.19

We used the GRADE framework as a starting point as it allows determination of research gaps based on the strength of the recommendation and the certainty of the evidence. For example, identification of a conditional recommendation, or low or very low certainty evidence, regardless of the strength of the recommendation, is indicative of where further research is required (see figure 1).

Previously, we had expanded the GRADE framework and included an additional component specific to the values and preferences of the community of users and potential users the recommendations intend to serve.12 We applied this step to the current guidelines on self- care interventions and reviewed results from the WHO- commissioned global online survey of users of sexual and reproductive self- care interventions.3 In addition, we captured specific research gaps identified by GDG participants during dedicated sessions at both the scoping and guidelines development meetings. We then viewed each research question through a GER lens and through discussion and consensus between authors of this paper identified the GER standards and principles

on July 28, 2020 by guest. Protected by copyright.

http://gh.bmj.com

/B

MJ G

lob Health: first published as 10.1136/bm

jgh-2019-002128 on 30 March 2020. D

ownloaded from

Page 4

4 Siegfried N, et al. BMJ Global Health 2020;5:e002128. doi:10.1136/bmjgh-2019-002128

BMJ Global Health

Figure 1 Hierarchical decision making algorithm to formulate research questions based on the presence and strength of a WHO recommendation, combined with the source of the question. GDG, Guidelines Development Group; GER, gender, equity and human rights; PICO, Population, Interventions, Comparison and Outcomes.

essential to consider and measure when planning future research.

Following the GDG meeting, for each a priori PICO question, and for additional research questions identi-fied from the global survey and during the meeting, we tabulated the following:1. The clinical, programmatic, values- based or GER re-

search question.2. The related recommendation formulated by the GDG

(or a record that no recommendation was made) where applicable.

3. The strength of each recommendation where applica-ble.

4. The certainty of the evidence underpinning the rec-ommendation where applicable.

5. The GER standards and principles most relevant to the question as discussed during the scoping and guide-line development meetings and agreed on by authors of this paper.

From the above, a research gap was identified when one or more of the following conditions was met: (1) the recommendation was conditional; or (2) the certainty of evidence was low or very low (even in the presence of a strong recommendation); or (3) key GER principles were not considered or absent (even if the certainty of evidence was moderate or high for clinical or public health effectiveness outcomes); or (4) no recommendation was made and the GDG or survey had articulated a research gap during their deliberations. We also identified a gap for strong recommendations if the evidence only arose

from well- resourced settings. For each identified research gap, the authors then tabulated the following:6. The most robust study design to answer the research

question(s), including identification of GER princi-ples to consider.

7. Conceptualisation of a pragmatic alternative study de-sign if the ideal design was not feasible.

8. Consideration of the methodological challenges of the alternative study design with the potential bias(es) identified.

Following tabulation, we categorised the primary focus of each research question according to the relevant GRADE domain as outlined in table 1.

Patient and public involvementParticipation of the public in developing research gaps was achieved in several ways: (1) through completion of the global survey, (2) members of several civil society associations contributed to the GDG meeting as partic-ipants on the GDG, (3) one of the authors is a patient representative and (4) all authors identify as individuals who engage or may engage with SRHR self- care interven-tions.

resulTsTables 2–6 provide an overview of research gaps in SRHR self- care, the most relevant GER for each question and the study design(s) considered most feasible and appro-priate to answer each question. Thirty- nine research ques-tions were formulated with seven overarching research

on July 28, 2020 by guest. Protected by copyright.

http://gh.bmj.com

/B

MJ G

lob Health: first published as 10.1136/bm

jgh-2019-002128 on 30 March 2020. D

ownloaded from

Page 5

Siegfried N, et al. BMJ Global Health 2020;5:e002128. doi:10.1136/bmjgh-2019-002128 5

BMJ Global Health

Tab

le 2

R

esea

rch

gap

s fo

r se

lf- ad

min

istr

atio

n of

inje

ctab

le c

ontr

acep

tion

Clin

ical

, pro

gra

mm

atic

, val

ues-

bas

ed o

r hu

man

rig

hts

que

stio

nC

urre

nt W

HO

re

com

men

dat

ion

Str

eng

th o

f re

com

men

dat

ion

So

urce

or

qua

lity

of

evid

ence

Key

GE

R

cons

ider

atio

nsId

eal s

tud

y d

esig

n(s)

Feas

ibili

ty a

nd p

ract

ical

co

nstr

aint

sA

lter

nati

ve s

tud

y d

esig

nM

etho

do

log

ical

issu

es a

risi

ng

in t

he a

lter

nati

ve s

tud

y d

esig

n

Ben

efits

ver

sus

harm

sP

ICO

: sho

uld

sel

f-

adm

inis

tere

d in

ject

able

co

ntra

cep

tion

be

mad

e av

aila

ble

as

an a

dd

ition

al

app

roac

h to

del

iver

in

ject

able

con

trac

eptio

n?

Sel

f- ad

min

iste

red

in

ject

able

con

trac

eptio

n sh

ould

be

mad

e av

aila

ble

as

an

add

ition

al a

pp

roac

h to

del

iver

inje

ctab

le

cont

race

ptio

n.

Str

ong.

Mod

erat

e.N

o d

ata

avai

lab

le

from

LM

IC;

avai

lab

ility

, ac

cess

ibili

ty,

priv

acy

and

co

nfid

entia

lity.

RC

T co

nduc

ted

in

LMIC

with

incl

usio

n of

ou

tcom

es t

o m

easu

re

acce

ssib

ility

, priv

acy

and

co

nfid

entia

lity.

As

a st

rong

WH

O

reco

mm

end

atio

n ex

ists

, it

wou

ld b

e un

ethi

cal t

o re

pea

t an

RC

T of

the

in

terv

entio

n.

Pro

spec

tive

non-

cont

rolle

d c

ohor

t st

udy

to b

e co

nduc

ted

in s

ever

al

low

- an

d m

idd

le- i

ncom

e co

untr

ies

whe

re t

he in

terv

entio

n is

or

will

be

imp

lem

ente

d. A

naly

sis

to c

omb

ine

qua

ntita

tive

surv

ey s

upp

lem

ente

d

with

focu

sed

in- d

epth

qua

litat

ive

inte

rvie

ws

with

wom

en (b

oth

thos

e w

ho h

ave

used

and

hav

e no

t us

ed s

elf-

adm

inis

tere

d in

ject

able

co

ntra

cep

tion

and

incl

udin

g ad

oles

cent

girl

s an

d y

oung

wom

en)

to e

xplo

re G

ER

out

com

es o

f av

aila

bili

ty, a

cces

sib

ility

, priv

acy

and

co

nfid

entia

lity.

Cos

t an

alys

es c

an b

e ne

sted

in t

he c

ohor

t st

udy.

Con

foun

din

g: p

artic

ipan

ts

in c

ohor

t se

lf- se

lect

to

self-

ad

min

iste

r in

ject

able

and

may

ha

ve o

ther

sim

ilar

char

acte

ristic

s th

at m

ay in

fluen

ce t

heir

abili

ty t

o ac

cess

inje

ctab

le c

ontr

acep

tion.

Mea

sur e

men

t b

ias:

sel

f- re

por

ting

bia

s an

issu

e fo

r q

ualit

ativ

e co

mp

onen

t. T

his

will

req

uire

w

ell t

rain

ed r

esea

rche

rs a

nd

valid

ated

too

ls. T

he t

riang

ulat

ion

of t

he q

ualit

ativ

e d

ata

with

th

e q

uant

itativ

e su

rvey

res

ults

p

rovi

des

an

elem

ent

of a

dd

ition

al

valid

atio

n of

out

com

es.

Valu

es a

nd

pre

fere

nces

Is t

here

an

imp

act

of

diff

eren

ces

bet

wee

n gr

oup

s of

wom

en (e

g,

age,

soc

ioec

onom

ic a

nd

occu

pat

iona

l sta

tus)

on

the

ir va

lues

and

p

refe

renc

es w

ith r

esp

ect

to s

elf-

adm

inis

trat

ion

of

inje

ctab

le c

ontr

acep

tion?

Not

ap

plic

able

.N

ot a

pp

licab

le.

GD

G.

Acc

essi

bili

ty,

priv

acy

and

co

nfid

entia

lity,

non

- d

iscr

imin

atio

n.

Cro

ss- s

ectio

nal s

tud

y of

w

omen

usi

ng in

ject

able

co

ntra

cep

tion

to

eval

uate

ass

ocia

tions

b

etw

een

char

acte

ristic

s an

d u

se, o

r w

illin

gnes

s to

use

.

A s

urve

y m

ay n

ot p

rovi

de

an o

bje

ctiv

e m

easu

re

or a

tru

e re

flect

ion

of

beh

avio

urs

if d

riven

by

que

stio

ns r

equi

ring

soci

ally

ac

cep

tab

le r

esp

onse

s.

This

is a

feas

ible

stu

dy

des

ign

but

co

uld

be

sup

ple

men

ted

with

dia

ry

colle

ctio

n of

dat

a as

gre

ater

priv

acy

and

less

like

ly fo

r so

cial

des

irab

ility

to

influ

ence

sel

f- re

cord

ing.

Par

tner

ing

with

wom

en w

ho h

ave

used

or

are

will

ing

to s

elf-

adm

inis

ter

inje

ctab

le

cont

race

ptio

n is

imp

orta

nt t

o id

entif

y cr

itica

l and

non

- jud

gmen

tal q

uest

ions

.

Mea

sure

men

t b

ias:

for

bot

h su

rvey

and

dia

ry c

olle

ctio

n,

que

stio

ns w

ill n

eed

to

be

form

ulat

ed t

oget

her

with

wom

en

usin

g in

ject

able

con

trac

eptio

n to

ens

ure

trus

t, p

artic

ipat

ion

and

va

lidity

to

avoi

d s

ocia

l des

irab

ility

b

ias.

Wha

t ha

pp

ens

afte

r w

omen

dis

cont

inue

us

e of

sel

f- in

ject

able

co

ntra

cep

tion

– d

o th

ey

use

othe

r m

etho

ds?

Not

ap

plic

able

.N

ot a

pp

licab

le.

GD

G.

Acc

essi

bili

ty.

Nat

iona

l pha

rmac

y or

cl

inic

- bas

ed r

egis

try

of w

omen

usi

ng s

elf-

ad

min

iste

red

inje

ctab

le

cont

race

ptio

n.

Reg

istr

ies

can

be

cost

ly

to e

stab

lish,

mai

ntai

n an

d

mon

itor.

A r

egis

try

req

uire

s p

oliti

cal w

ill a

nd c

apac

ity

to p

rovi

de

effe

ctiv

e d

ata

to

info

rm d

ecis

ions

.

Con

veni

ence

sam

ple

and

sur

vey

of

wom

en a

tten

din

g fa

mily

pla

nnin

g cl

inic

s to

elu

cid

ate

thei

r p

rior

mod

es

of c

ontr

acep

tion

and

rea

sons

for

dis

cont

inua

tion.

Sel

ectio

n b

ias:

suc

h a

sam

ple

w

ould

onl

y in

clud

e w

omen

who

at

tend

the

clin

ic t

o ch

ange

to

a d

iffer

ent

form

of c

ontr

acep

tion;

w

omen

who

cea

se t

o us

e se

lf-

adm

inis

tere

d c

ontr

acep

tion

and

d

o no

t re

turn

to

the

clin

ic w

ould

no

t b

e in

clud

ed. T

heir

choi

ces

may

diff

er fr

om t

hose

wom

en

who

do

atte

nd t

he c

linic

.

Wha

t is

the

imp

act

of

stig

ma

on t

he c

hoic

e of

se

lf- ad

min

istr

atio

n of

in

ject

able

con

trac

eptio

n?

Not

ap

plic

able

.N

ot a

pp

licab

le.

GD

G a

nd

surv

ey.

Acc

essi

bili

ty,

priv

acy

and

co

nfid

entia

lity

and

no

n- d

iscr

imin

atio

n.

Qua

litat

ive

stud

y of

in

terv

iew

s of

wom

en

self-

adm

inis

terin

g in

ject

able

con

trac

eptio

n to

elu

cid

ate

thei

r ex

per

ienc

es a

nd

obse

rvat

ions

.

This

is a

feas

ible

stu

dy.

Not

ap

plic

able

.S

elec

tion

bia

s: a

s w

ith a

ll q

ualit

ativ

e st

udie

s, p

artic

ipan

ts

are

pur

pos

ivel

y se

lect

ed.

How

ever

, thi

s is

an

adva

ntag

e in

thi

s st

udy

des

ign

as t

hose

w

omen

who

hav

e co

nsid

ered

thi

s is

sue

will

hav

e ric

h ex

per

ienc

es

and

ob

serv

atio

ns t

o sh

are.

Wha

t ar

e th

e op

timal

m

odel

s of

info

rmat

ion

pro

visi

on fo

r aw

aren

ess

rais

ing

and

incr

easi

ng

know

led

ge o

f sel

f-

adm

inis

trat

ion

of

inje

ctab

le c

ontr

acep

tion?

Not

ap

plic

able

.N

ot a

pp

licab

le.

GD

G a

nd

surv

ey.

Acc

essi

bili

ty, t

he

right

to

seek

, re

ceiv

e an

d im

par

t in

form

atio

n.

Com

par

ativ

e ef

fect

iven

ess

rese

arch

of

diff

eren

t m

odel

s of

in

form

atio

n p

rovi

sion

, op

timal

ly in

an

RC

T.

This

will

firs

t re

qui

re

dev

elop

men

t of

mod

els

of d

edic

ated

info

rmat

ion

pro

visi

on a

nd h

ealth

lit

erac

y (e

g, t

elep

honi

c,

inte

rnet

, pos

ters

and

co

unse

lling

) prio

r to

ex

per

imen

tal t

estin

g.

Exp

lora

tory

qua

litat

ive

inte

rvie

ws

with

wom

en s

elf-

adm

inis

terin

g or

will

ing

to s

elf-

adm

inis

ter

to

iden

tify

thei

r kn

owle

dge

, att

itud

es

and

und

erst

and

ing

of in

ject

able

co

ntra

cep

tion

are

req

uire

d t

o in

form

m

odel

s of

info

rmat

ion

pro

visi

on.

Sel

ectio

n b

ias:

as

with

all

qua

litat

ive

stud

ies,

par

ticip

ants

ar

e p

urp

osiv

ely

sele

cted

. H

owev

er, t

his

is a

n ad

vant

age

in t

his

stud

y d

esig

n as

tho

se

wom

en w

ho h

ave

cons

ider

ed t

his

issu

e w

ill h

ave

rich

exp

erie

nces

an

d o

bse

rvat

ions

to

shar

e w

hich

ca

n gu

ide

the

dev

elop

men

t of

he

alth

pro

mot

ion

inte

rven

tion

and

too

ls.

Con

tinue

d

on July 28, 2020 by guest. Protected by copyright.

http://gh.bmj.com

/B

MJ G

lob Health: first published as 10.1136/bm

jgh-2019-002128 on 30 March 2020. D

ownloaded from

Page 6

6 Siegfried N, et al. BMJ Global Health 2020;5:e002128. doi:10.1136/bmjgh-2019-002128

BMJ Global Health

Clin

ical

, pro

gra

mm

atic

, val

ues-

bas

ed o

r hu

man

rig

hts

que

stio

nC

urre

nt W

HO

re

com

men

dat

ion

Str

eng

th o

f re

com

men

dat

ion

So

urce

or

qua

lity

of

evid

ence

Key

GE

R

cons

ider

atio

nsId

eal s

tud

y d

esig

n(s)

Feas

ibili

ty a

nd p

ract

ical

co

nstr

aint

sA

lter

nati

ve s

tud

y d

esig

nM

etho

do

log

ical

issu

es a

risi

ng

in t

he a

lter

nati

ve s

tud

y d

esig

n

Hum

an r

ight

s an

d e

qui

tyW

hat

imp

lem

enta

tion

mea

sure

s ca

n en

sure

th

at in

equi

ty is

red

uced

or

min

imis

ed w

hen

self-

adm

inis

trat

ion

is

intr

oduc

ed?

Not

ap

plic

able

.N

ot a

pp

licab

le.

GD

G.

Non

- dis

crim

inat

ion.

Com

par

ativ

e ef

fect

iven

ess

rese

arch

of

diff

eren

t m

odel

s of

imp

lem

enta

tion,

op

timal

ly in

a c

lust

er

RC

T.

An

RC

T of

mod

els

of

imp

lem

enta

tion

may

b

e ve

ry c

ostly

due

to

the

com

ple

x na

ture

of

imp

lem

enta

tion

stra

tegi

es

that

are

dep

end

ent

on t

he

sett

ing,

typ

e of

pro

vid

er

and

tra

inin

g re

qui

red

, and

en

surin

g fid

elity

to

the

inte

rven

tion.

Des

crip

tive

case

stu

die

s of

d

emon

stra

tion

pro

ject

s us

ing

diff

eren

t st

rate

gies

to

imp

lem

ent

inje

ctab

le

cont

race

ptio

n in

pro

gram

mes

in

seve

ral c

ount

ries

to in

dic

ate

optim

al

mea

sure

s to

ens

ure

equi

tab

le a

cces

s an

d t

o fo

rmul

ate

less

ons

lear

nt fo

r sc

ale-

up e

lsew

here

Lack

of g

ener

alis

abili

ty: c

urre

nt

pra

ctic

es a

nd p

oten

tial f

or

scal

e w

ill d

iffer

mar

ked

ly

bet

wee

n co

untr

ies.

The

refo

re

sele

ctio

n of

cou

ntry

site

s fo

r d

emon

stra

tion

pro

ject

s w

ill

req

uire

cha

ract

eris

atio

n of

un

der

lyin

g re

sour

ce n

eed

s,

trai

ning

, sta

ff an

d s

ocio

pol

itica

l fa

ctor

s to

ens

ure

max

imal

d

iver

sity

and

tha

t re

ason

able

co

mp

aris

ons

can

be

mad

e b

etw

een

coun

trie

s. T

riang

ulat

ion

bet

wee

n d

ata

sour

ces

and

site

s w

ill b

e ne

cess

ary

to in

crea

se

gene

ralis

abili

ty.

Acc

epta

bili

tyIs

the

re a

n im

pac

t of

d

iffer

ence

s b

etw

een

heal

thca

re p

rovi

der

s (e

g, a

ge, i

ncom

e st

atus

of

cou

ntry

, priv

ate/

pub

lic s

ecto

r) o

n th

e ac

cep

tab

ility

of

self-

adm

inis

trat

ion

of

inje

ctab

le c

ontr

acep

tion?

Not

ap

plic

able

.N

ot a

pp

licab

le.

GD

G.

Acc

epta

bili

ty a

nd

acce

ssib

ilty.

Cro

ss- s

ectio

nal s

tud

y of

hea

lthca

re p

rovi

der

s to

eva

luat

e as

soci

atio

ns

bet

wee

n ch

arac

teris

tics

and

will

ingn

ess

to

pre

scrib

e an

d p

rovi

de

inje

ctab

le c

ontr

acep

tion

for

self-

adm

inis

trat

ion.

A s

urve

y m

ay n

ot p

rovi

de

an o

bje

ctiv

e m

easu

re

or a

tru

e re

flect

ion

of

beh

avio

urs

if d

riven

by

que

stio

ns r

equi

ring

soci

ally

ac

cep

tab

le r

esp

onse

s.

This

is a

feas

ible

stu

dy

des

ign

but

co

uld

be

sup

ple

men

ted

with

focu

s gr

oup

s or

qua

litat

ive

inte

rvie

ws.

Mea

sure

men

t b

ias:

for

bot

h su

rvey

and

inte

rvie

ws,

que

stio

ns

will

nee

d t

o b

e fo

rmul

ated

to

geth

er w

ith h

ealth

care

p

rovi

der

s to

ens

ure

trus

t,

par

ticip

atio

n an

d v

alid

ity a

nd t

o av

oid

soc

ial d

esira

bili

ty b

ias.

Wha

t is

the

sca

le

and

con

seq

uenc

e of

in

corr

ect

use

of s

elf-

ad

min

istr

atio

n?

Not

ap

plic

able

.N

ot a

pp

licab

le.

GD

G.

Acc

ount

abili

ty.

Nat

iona

l pha

rmac

y or

cl

inic

- bas

ed r

egis

try

of w

omen

usi

ng s

elf-

ad

min

iste

red

inje

ctab

le

cont

race

ptio

n.

Reg

istr

ies

can

be

cost

ly

to e

stab

lish,

mai

ntai

n an

d

mon

itor.

A r

egis

try

req

uire

s p

oliti

cal w

ill a

nd c

apac

ity

to p

rovi

de

effe

ctiv

e d

ata

to

info

rm d

ecis

ions

.

Sen

tinel

(act

ive)

sur

veill

ance

thr

ough

m

onito

ring

and

eva

luat

ion

of a

dve

rse

effe

cts

and

ad

vers

e ev

ents

at

clin

ic

or p

rovi

der

leve

l em

bed

ded

in lo

cal

and

nat

iona

l qua

lity

assu

ranc

e p

rogr

amm

es.

Rep

ortin

g b

ias:

the

dat

a co

llect

ed

is d

epen

den

t on

the

tra

inin

g an

d e

xper

tise

of p

rovi

der

s to

ad

equa

tely

rec

ogni

se a

dve

rse

even

ts. R

esou

rces

can

be

dire

cted

tow

ard

s ap

pro

pria

te

sent

inel

site

s to

ens

ure

staf

f are

sk

illed

and

tha

t d

ata

inte

grity

is

ensu

red

.

Tab

le 2

C

ontin

ued

Con

tinue

d

on July 28, 2020 by guest. Protected by copyright.

http://gh.bmj.com

/B

MJ G

lob Health: first published as 10.1136/bm

jgh-2019-002128 on 30 March 2020. D

ownloaded from

Page 7

Siegfried N, et al. BMJ Global Health 2020;5:e002128. doi:10.1136/bmjgh-2019-002128 7

BMJ Global Health

Clin

ical

, pro

gra

mm

atic

, val

ues-

bas

ed o

r hu

man

rig

hts

que

stio

nC

urre

nt W

HO

re

com

men

dat

ion

Str

eng

th o

f re

com

men

dat

ion

So

urce

or

qua

lity

of

evid

ence

Key

GE

R

cons

ider

atio

nsId

eal s

tud

y d

esig

n(s)

Feas

ibili

ty a

nd p

ract

ical

co

nstr

aint

sA

lter

nati

ve s

tud

y d

esig

nM

etho

do

log

ical

issu

es a

risi

ng

in t

he a

lter

nati

ve s

tud

y d

esig

n

Res

ourc

e us

eW

hat

are

the

user

an

d p

rovi

der

cos

ts o

f se

lf- ad

min

istr

atio

n (a

s w

ell a

s ou

t- of

- poc

ket

exp

end

iture

s), c

omp

ared

w

ith p

rovi

der

- bas

ed

adm

inis

trat

ion?

Not

ap

plic

able

.N

ot a

pp

licab

le.

GD

G a

nd

surv

ey.

Acc

essi

bili

ty.

Em

piri

cal c

ostin

g st

udie

s to

est

imat

e us

er

and

pro

vid

er c

osts

, an

d fi

nanc

ing

sour

ces

for

each

(use

rs p

ay o

r he

alth

sys

tem

pay

s)D

ata

are

curr

ently

onl

y av

aila

ble

from

one

st

udy

in t

hree

cou

ntrie

s in

sub

- Sah

aran

Afr

ica.

G

iven

tha

t co

sts

vary

by

cont

ext

and

by

del

iver

y m

odel

, the

r e is

a n

eed

fo

r m

ore

rep

rese

ntat

ive

cost

est

imat

es t

o co

nfirm

tha

t us

er c

osts

an

d o

ut- o

f- p

ocke

t ex

pen

ditu

res

are

likel

y to

dec

reas

e w

ith s

elf-

ad

min

istr

atio

n.

This

is a

feas

ible

stu

dy

but

ca

n b

e co

stly

to

cond

uct

acr o

ss a

rep

rese

ntat

ive

sam

ple

of u

sers

and

fa

cilit

ies.

This

is fe

asib

le b

ut c

ould

be

sup

ple

men

ted

by

mod

ellin

g b

ased

on

the

exis

ting

emp

irica

l cos

t d

ata,

cos

t st

ruct

ures

and

exp

ert

cons

ulta

tion

on

exp

ecte

d v

aria

tion

in p

aram

eter

s.

Mea

sure

men

t b

ias:

the

res

ults

ar

e d

epen

den

t on

use

r re

call

and

sel

f- r e

por

ted

res

ourc

e us

e,

opp

ortu

nity

cos

ts, a

s w

ell a

s on

as

sum

ptio

ns r

egar

din

g sh

ared

co

sts

in h

ealth

faci

litie

s.

Wha

t is

the

cos

t-

effe

ctiv

enes

s of

sel

f-

adm

inis

trat

ion,

com

par

ed

with

pro

vid

er- b

ased

ad

min

istr

atio

n?

Not

ap

plic

able

.N

ot a

pp

licab

le.

GD

G.

Acc

essi

bili

ty.

Mod

ellin

g st

udie

s of

co

st-

effe

ctiv

enes

s ac

ross

bro

ader

ran

ge o

f se

ttin

gs, t

akin

g b

oth

a he

alth

care

pro

vid

er a

nd

soci

etal

per

spec

tive.

This

is a

feas

ible

stu

dy

but

its

qua

lity

and

use

fuln

ess

dep

end

s on

the

rel

iab

ility

of

the

par

amet

ers

for

unit

cost

s an

d e

ffect

iven

ess,

w

hich

wou

ld b

e d

raw

n fr

om c

ostin

g st

udie

s (n

oted

ab

ove)

and

RC

Ts. I

t w

ill

also

req

uire

mor

e lo

ng-

term

mod

ellin

g to

cap

tur e

d

owns

trea

m e

ffect

s on

the

ty

pes

of c

ontr

acep

tives

se

lf- in

ject

ing

wom

en

switc

h to

com

par

ed w

ith

pro

vid

er- b

ased

inje

ctin

g w

omen

.

Mea

sure

men

t b

ias:

the

res

ults

ar

e d

epen

den

t on

sel

ectio

n of

var

iab

les

incl

uded

in t

he

mod

el, a

ssum

ptio

ns r

egar

din

g d

isco

unt

rate

s, c

ost-

effe

ctiv

enes

s th

resh

old

s an

d a

ccur

ate

estim

ates

of c

urre

nt a

nd

pro

ject

ed c

osts

and

effe

cts.

Wha

t is

the

en

viro

nmen

tal i

mp

act

of d

isp

osal

of s

elf-

ad

min

iste

r ed

inje

ctab

le

oral

con

trac

eptio

n?

Not

ap

plic

able

.N

ot a

pp

licab

le.

GD

G.

Acc

ount

abili

ty.

Col

lect

ion

and

col

latio

n of

rep

orts

from

en

viro

nmen

tal r

egul

ator

y b

odie

s d

ocum

entin

g sc

ale

and

imp

act

of

med

ical

was

te a

nd

hous

ehol

d w

aste

.

Reg

ulat

ory

agen

cies

are

un

likel

y to

diff

eren

tiate

b

etw

een

whe

ther

inje

ctab

le

cont

race

ptiv

es p

rese

nt in

w

aste

are

sou

rced

dire

ctly

fr

om c

linic

s or

from

the

ho

mes

of u

sers

.

A q

ualit

ativ

e ho

useh

old

dia

ry

stud

y of

wom

en w

ho s

elf-

inje

ct in

w

hich

the

y re

cord

was

te fr

om t

he

pac

kagi

ng a

nd t

heir

man

agem

ent

of

it. T

his

coul

d b

e d

eliv

ered

with

the

in

ject

able

pac

kagi

ng a

nd w

omen

can

re

turn

it w

hen

colle

ctin

g th

eir

next

p

resc

riptio

n.

Rec

all b

ias:

wom

en m

ay n

ot

rem

emb

er w

hat

they

did

with

th

e p

acka

ging

if t

hey

do

not

reco

rd it

at

the

time.

Thi

s m

ay

req

uire

an

ince

ntiv

e d

urin

g th

e re

sear

ch p

erio

d in

ord

er t

o en

cour

age

wom

en t

o co

mp

lete

th

e ho

useh

old

was

te d

iary

.

GD

G, G

uid

elin

e D

evel

opm

ent

Gro

up; G

ER

, gen

der

, eq

uity

and

hum

an r

ight

s; R

CTs

, ran

dom

ised

con

trol

led

tria

ls.

Tab

le 2

C

ontin

ued

on July 28, 2020 by guest. Protected by copyright.

http://gh.bmj.com

/B

MJ G

lob Health: first published as 10.1136/bm

jgh-2019-002128 on 30 March 2020. D

ownloaded from

Page 8

8 Siegfried N, et al. BMJ Global Health 2020;5:e002128. doi:10.1136/bmjgh-2019-002128

BMJ Global Health

Tab

le 3

R

esea

rch

gap

s fo

r se

lf- co

llect

ion

of S

TI s

amp

les

Clin

ical

, pro

gra

mm

atic

, val

ues-

b

ased

or

hum

an r

ight

s q

uest

ion

Cur

rent

WH

O

reco

mm

end

atio

nS

tren

gth

of

reco

mm

end

atio

n

So

urce

or

qua

lity

of

evid

ence

Key

GE

R

cons

ider

atio

nsId

eal s

tud

y d

esig

n(s)

Feas

ibili

ty

and

pra

ctic

al

cons

trai

nts

Alt

erna

tive

stu

dy

des

ign

Met

hod

olo

gic

al is

sues

ari

sing

in

the

alt

erna

tive

stu

dy

des

ign

Ben

efits

ve

rsus

har

ms

PIC

O: s

houl

d

self-

colle

ctio

n of

sa

mp

les

for

STI

s b

e m

ade

avai

lab

le

as a

n ad

diti

onal

ap

pro

ach

to

del

iver

STI

tes

ting

serv

ices

?

Sel

f- co

llect

ion

of

sam

ple

s fo

r N

eiss

eria

go

norr

hoea

e an

d

Chl

amyd

ia t

rach

omat

is

shou

ld b

e m

ade

avai

lab

le a

s an

ad

diti

onal

ap

pro

ach

to d

eliv

er S

TI t

estin

g se

rvic

es.

Str

ong.

Mod

erat

e.N

o d

ata

from

LM

IC. A

vaila

bili

ty

and

acc

essi

bili

ty.

RC

T co

nduc

ted

in

LM

IC w

ith

incl

usio

n of

ou

tcom

es

to m

easu

re

avai

lab

ility

and

ac

cess

ibili

ty.

As

a st

rong

WH

O

reco

mm

end

atio

n ex

ists

, it

wou

ld b

e un

ethi

cal t

o re

pea

t an

RC

T of

the

in

terv

entio

n.

Pro

spec

tive

non-

co

ntro

lled

coh

ort

stud

y to

be

cond

ucte

d in

se

vera

l LM

ICs

whe

re t

he

inte

rven

tion

is o

r w

ill b

e im

ple

men

ted

. Ana

lysi

s to

com

bin

e q

uant

itativ

e su

rvey

sup

ple

men

ted

w

ith fo

cuse

d in

- dep

th

qua

litat

ive

inte

rvie

ws

with

gen

eral

pop

ulat

ion

(with

att

entio

n gi

ven

to e

nsur

e sa

mp

ling

of

vuln

erab

le p

opul

atio

ns)

to e

xplo

re G

ER

ou

tcom

es o

f ava

ilab

ility

an

d a

cces

sib

ility

. Cos

t an

alys

es c

an b

e ne

sted

in

the

coho

rt s

tud

y.

Con

foun

din

g: p

artic

ipan

ts

in c

ohor

t se

lf- se

lect

to

self-

co

llect

and

may

hav

e ot

her

sim

ilar

char

acte

ristic

s th

at m

ay

influ

ence

the

ir ab

ility

to

acce

ss

STI

sel

f- co

llect

ion.

Mea

sure

men

t b

ias:

sel

f-

rep

ortin

g b

ias

an is

sue

for

qua

litat

ive

com

pon

ent.

Thi

s w

ill

req

uire

wel

l- tr

aine

d r

esea

rche

rs

and

val

idat

ed t

ools

. The

tr

iang

ulat

ion

of t

he q

ualit

ativ

e d

ata

with

the

qua

ntita

tive

surv

ey r

esul

ts p

rovi

des

an

elem

ent

of a

dd

ition

al v

alid

atio

n of

out

com

es.

Sel

f- co

llect

ion

of

sam

ple

s fo

r Tre

pon

ema

pal

lidum

(syp

hilis

) and

Tr

icho

mon

as v

agin

alis

m

ay b

e co

nsid

ered

as

an a

dd

ition

al a

pp

roac

h to

del

iver

STI

tes

ting

serv

ices

.

Con

diti

onal

.Lo

w.

Ava

ilab

ility

, ac

cess

ibili

ty a

nd

acce

pta

bili

ty.

Ran

dom

ised

co

ntro

lled

tria

l sp

ecifi

c to

se

lf- co

llect

ion

of t

hese

or

gani

sms.

This

is a

feas

ible

d

esig

n.

G

ener

alis

abili

ty: t

he c

hoic

e of

se

ttin

g an

d u

ser

pop

ulat

ion

will

det

erm

ine

how

wid

ely

the

resu

lts c

an b

e ap

plie

d. T

he

focu

s sh

ould

be

on t

estin

g th

is in

LM

IC s

ettin

gs a

nd w

ith

vuln

erab

le p

opul

atio

ns.

Wha

t is

the

imp

act

of s

elf-

sam

plin

g fo

r S

TIs

on p

artn

er

scre

enin

g?

GD

G.

Priv

acy

and

co

nfid

entia

lity.

Ran

dom

ised

co

ntro

lled

tria

l w

ith u

pta

ke

of p

artn

er

scre

enin

g in

clud

ed a

s an

ou

tcom

e.

As

a st

rong

WH

O

reco

mm

end

atio

n ex

ists

, it

wou

ld b

e un

ethi

cal t

o re

pea

t an

RC

T of

the

in

terv

entio

n.

Cro

ss- s

ectio

nal s

tud

y of

w

omen

who

res

pon

d t

o re

sults

follo

win

g S

TI s

elf-

co

llect

ion

to a

scer

tain

how

m

any

have

not

ified

the

ir p

artn

ers

to b

e sc

r een

ed,

and

of t

hose

how

man

y p

artn

ers

have

gon

e fo

r sc

reen

ing.

Acc

epta

bili

ty

of s

cree

ning

for

par

tner

s,

and

ava

ilab

ility

and

ac

cess

ibili

ty o

f par

tner

sc

reen

ing

will

be

imp

orta

nt

to m

easu

re.

Sel

ectio

n b

ias:

sur

veys

will

on

ly c

aptu

re t

hose

wom

en

who

hav

e re

spon

ded

act

ivel

y to

the

res

ults

and

may

not

be

rep

rese

ntat

ive

of t

hose

wom

en

who

do

not

resp

ond

act

ivel

y to

th

eir

resu

lts.

Con

tinue

d

on July 28, 2020 by guest. Protected by copyright.

http://gh.bmj.com

/B

MJ G

lob Health: first published as 10.1136/bm

jgh-2019-002128 on 30 March 2020. D

ownloaded from

Page 9

Siegfried N, et al. BMJ Global Health 2020;5:e002128. doi:10.1136/bmjgh-2019-002128 9

BMJ Global Health

Clin

ical

, pro

gra

mm

atic

, val

ues-

b

ased

or

hum

an r

ight

s q

uest

ion

Cur

rent

WH

O

reco

mm

end

atio

nS

tren

gth

of

reco

mm

end

atio

n

So

urce

or

qua

lity

of

evid

ence

Key

GE

R

cons

ider

atio

nsId

eal s

tud

y d

esig

n(s)

Feas

ibili

ty

and

pra

ctic

al

cons

trai

nts

Alt

erna

tive

stu

dy

des

ign

Met

hod

olo

gic

al is

sues

ari

sing

in

the

alt

erna

tive

stu

dy

des

ign

Wha

t is

the

imp

act

of s

elf-

sam

plin

g fo

r S

TIs

on li

nkag

e to

car

e an

d c

ase-

fin

din

g?

GD

G.

Acc

epta

bili

ty

and

non

- d

iscr

imin

atio

n.

RC

T w

ith

incl

usio

n of

ou

tcom

es

to m

easu

r e

linka

ge t

o ca

re

and

cas

e-

find

ing.

As

a st

rong

WH

O

reco

mm

end

atio

n ex

ists

, it

wou

ld

be

unet

hica

l to

rep

eat

an R

CT

of

effe

ctiv

enes

s of

the

in

terv

entio

n.

Inte

rrup

ted

tim

e se

ries

usin

g re

pea

ted

cro

ss-

sect

iona

l sur

veys

of

pop

ulat

ions

att

end

ing

STI

clin

ics

for

trea

tmen

t b

efor

e, d

urin

g an

d a

fter

im

ple

men

tatio

n of

pol

icy

to a

scer

tain

pro

por

tion

of

wom

en w

ho s

elf-

sam

ple

d

prio

r to

link

age

to c

are.

Att

ritio

n b

ias:

the

res

ults

may

in

dic

ate

that

mor

e p

eop

le w

ho

link

to c

are

used

sel

f- sa

mp

ling,

b

ut t

his

doe

s no

t p

r ovi

de

info

rmat

ion

on h

ow m

any

peo

ple

who

sel

f- sa

mp

led

with

a

pos

itive

res

ult

did

not

link

to

car

e.

Wha

t is

the

ben

efit

and

har

m o

f sel

f-

sam

plin

g fo

r S

TIs

of

vira

l aet

iolo

gy?

GD

G.

Ava

ilab

ility

, ac

cess

ibili

ty a

nd

acce

pta

bili

ty.

Ran

dom

ised

co

ntro

lled

tria

l sp

ecifi

c to

se

lf- co

llect

ion

of S

TIs

of

vira

l aet

iolo

gy.

Out

com

es

to in

clud

e m

easu

rem

ent

of a

vaila

bili

ty

and

ac

cess

ibili

ty,

as w

ell a

s ac

cep

tab

ility

of

self-

sam

plin

g,

esp

ecia

lly

rega

rdin

g sa

mp

ling

of

blo

ods

(eg,

he

pat

itis

B

viru

s).

This

is a

feas

ible

d

esig

n.

P

erfo

rman

ce b

ias:

it is

not

p

ossi

ble

to

mas

k us

ers

or

pro

vid

ers

to m

ode

of s

amp

ling.

Th

is c

an b

e m

itiga

ted

by

cons

ider

ing

cond

uctin

g a

clus

ter

RC

T ra

ther

tha

n an

in

div

idua

lly r

and

omis

ed t

rial,

or

a st

epp

ed w

edge

des

ign

whe

re

the

inte

rven

tion

is d

eliv

ered

at

stag

gere

d t

ime

per

iod

s.

Valu

es a

nd

pre

fere

nces

Wha

t ar

e th

e va

lues

an

d p

refe

renc

es

of m

argi

nalis

ed

pop

ulat

ions

(eg,

M

en w

ho h

ave

sex

with

men

, se

x w

orke

rs,

tran

s p

opul

atio

ns)

rega

rdin

g se

lf-

sam

plin

g?

Not

ap

plic

able

.

G

DG

and

su

rvey

.A

cces

sib

ility

ac

cep

tab

ility

, p

rivac

y an

d

confi

den

tialit

y,

and

non

- d

iscr

imin

atio

n.

Qua

litat

ive

key

info

rman

t st

udy

of w

omen

and

m

en w

ho a

re

mar

gina

lised

re

gard

ing

thei

r va

lues

and

p

refe

renc

es o

f se

lf- sa

mp

ling

and

the

bar

riers

th

at t

hey

exp

erie

nce

or

may

exp

erie

nce

whe

n ac

cess

ing

self-

sam

plin

g,

rece

ivin

g th

eir

resu

lts a

nd

whe

n lin

king

to

care

.

This

is a

feas

ible

st

udy.

Sel

ectio

n b

ias:

as

with

al

l qua

litat

ive

stud

ies,

p

artic

ipan

ts a

re p

urp

osiv

ely

sele

cted

. How

ever

, thi

s is

an

ad

vant

age

in t

his

stud

y d

esig

n as

mar

gina

lised

w

omen

and

men

—in

clud

ing

tran

s p

opul

atio

ns—

who

hav

e co

nsid

ered

thi

s is

sue

will

ha

ve r

ich

exp

erie

nces

and

ob

serv

atio

ns t

o sh

are.

Tab

le 3

C

ontin

ued

Con

tinue

d

on July 28, 2020 by guest. Protected by copyright.

http://gh.bmj.com

/B

MJ G

lob Health: first published as 10.1136/bm

jgh-2019-002128 on 30 March 2020. D

ownloaded from

Page 10

10 Siegfried N, et al. BMJ Global Health 2020;5:e002128. doi:10.1136/bmjgh-2019-002128

BMJ Global Health

Clin

ical

, pro

gra

mm

atic

, val

ues-

b

ased

or

hum

an r

ight

s q

uest

ion

Cur

rent

WH

O

reco

mm

end

atio

nS

tren

gth

of

reco

mm

end

atio

n

So

urce

or

qua

lity

of

evid

ence

Key

GE

R

cons

ider

atio

nsId

eal s

tud

y d

esig

n(s)

Feas

ibili

ty

and

pra

ctic

al

cons

trai

nts

Alt

erna

tive

stu

dy

des

ign

Met

hod

olo

gic

al is

sues

ari

sing

in

the

alt

erna

tive

stu

dy

des

ign

Hum

an r

ight

s an

d e

qui

tyW

hat

are

the

bes

t p

ract

ices

for

avoi

din

g co

erci

on in

se

lf- sa

mp

ling?

Not

ap

plic

able

.

G

DG

.N

on-

dis

crim

inat

ion,

p

rivac

y an

d

confi

den

tialit

y.

Qua

litat

ive

stud

y us

ing

focu

s gr

oup

s of

vul

nera

ble

p

opul

atio

ns in

co

ntex

ts w

here

se

lf- sa

mp

ling

is im

ple

men

ted

to

und

erst

and

ho

w t

hey

per

ceiv

e co

erci

on c

ould

b

e av

oid

ed.

This

can

allo

w

crea

tion

of

an e

xpan

ded

ve

rsio

n of

the

in

terv

entio

n fo

r fu

rthe

r ev

alua

tion

with

m

easu

rem

ent

of n

on- c

oerc

ion

incl

uded

as

an

outc

ome.

This

is a

feas

ible

d

esig

n.

S

elec

tion

bia

s: a

s w

ith a

ll q

ualit

ativ

e st

udie

s, p

artic

ipan

ts

are

pur

pos

ivel

y se

lect

ed.

How

ever

, thi

s is

an

adva

ntag

e in

thi

s st

udy

des

ign

as

mar

gina

lised

wom

en a

nd m

en

who

hav

e co

nsid

ered

thi

s is

sue

will

hav

e ric

h ex

per

ienc

es a

nd

obse

rvat

ions

to

shar

e.

Res

ourc

e us

eIs

sel

f- co

llect

ion

for

STI

s co

st- e

ffect

ive?

Not

ap

plic

able

.N

ot a

pp

licab

le.

GD

G.

Acc

essi

bili

ty.

Mod

ellin

g st

udie

s of

cos

t-

effe

ctiv

enes

s ac

ross

b

road

er r

ange

of

set

tings

, ta

king

bot

h a

heal

thca

re

pro

vid

er

and

soc

ieta

l p

ersp

ectiv

e.

This

is a

feas

ible

st

udy,

but

its

qua

lity

and

us

eful

ness

dep

end

on

the

rel

iab

ility

of

the

par

amet

ers

for

unit

cost

s an

d

effe

ctiv

enes

s,

whi

ch w

ould

b

e d

raw

n fr

om

cost

ing

stud

ies

and

the

RC

Ts

with

effe

ctiv

enes

s d

ata.

It w

ill a

lso

req

uire

mor

e lo

ng-

term

mod

ellin

g to

cap

ture

d

owns

trea

m

effe

cts

on li

nkag

e to

car

e.

Mea

sure

men

t b

ias:

the

re

sults

are

dep

end

ent

on

sele

ctio

n of

var

iab

les

incl

uded

in

the

mod

el, a

ssum

ptio

ns

rega

rdin

g d

isco

unt

rate

s,

cost

- effe

ctiv

enes

s th

resh

old

s an

d a

ccur

ate

estim

ates

of

curr

ent

and

pro

ject

ed c

osts

. Fo

r ex

amp

le, v

ulne

rab

le

pop

ulat

ions

with

lim

ited

acc

ess

to c

are

are

likel

y to

gen

erat

e d

iffer

ent

resu

lts b

ased

on

the

pro

bab

ility

tha

t th

ey w

ould

hav

e lim

ited

or

no a

cces

s to

car

e w

ithou

t se

lf- sa

mp

ling.

Dat

a ac

cura

cy: u

nit

cost

s m

ay n

ot b

e av

aila

ble

for

man

y or

any

LM

IC, e

spec

ially

w

here

sel

f- sa

mp

ling

has

not

bee

n av

aila

ble

. To

avoi

d

assu

mp

tions

, thi

s m

ay r

equi

re

prio

r as

sess

men

t vi

a an

initi

al

imp

lem

enta

tion

pro

cess

to

asce

rtai

n re

ason

able

cos

tings

.

Tab

le 3

C

ontin

ued

Con

tinue

d

on July 28, 2020 by guest. Protected by copyright.

http://gh.bmj.com

/B

MJ G

lob Health: first published as 10.1136/bm

jgh-2019-002128 on 30 March 2020. D

ownloaded from

Page 11

Siegfried N, et al. BMJ Global Health 2020;5:e002128. doi:10.1136/bmjgh-2019-002128 11

BMJ Global Health

Clin

ical

, pro

gra

mm

atic

, val

ues-

b

ased

or

hum

an r

ight

s q

uest

ion

Cur

rent

WH

O

reco

mm

end

atio

nS

tren

gth

of

reco

mm

end

atio

n

So

urce

or

qua

lity

of

evid

ence

Key

GE

R

cons

ider

atio

nsId

eal s

tud

y d

esig

n(s)

Feas

ibili

ty

and

pra

ctic

al

cons

trai

nts

Alt

erna

tive

stu

dy

des

ign

Met

hod

olo

gic

al is

sues

ari

sing

in

the

alt

erna

tive

stu

dy

des

ign

Doe

s se

lf- co

llect

ion

offe

r a

cost

- effi

cien

t ap

pro

ach

to c

ase

det

ectio

n in

low

- in

com

e se

ttin

gs?

Not

ap

plic

able

.N

ot a

pp

licab

le.

GD

G.

Ava

ilab

ility

and

ac

cess

ibili

ty.

Em

piri

cal

cost

ing

stud

ies

to e

stim

ate

the

cost

s an

d t

arge

ting

effic

ienc

y of

se

lf- co

llect

ion

in L

MIC

s us

ing

vario

us fe

asib

le

mod

els

of

del

iver

y.

As

abov

e.

A

s ab

ove.

GD

G, G

uid

elin

e D

evel

opm

ent

Gro

up; G

ER

, gen

der

, eq

uity

and

hum

an r

ight

s; P

ICO

, Pop

ulat

ion,

Inte

rven

tions

, Com

par

ison

and

Out

com

es; R

CTs

, ran

dom

ised

con

trol

led

tria

ls; S

TI, s

exua

lly t

rans

mitt

ed in

fect

ion.

Tab

le 3

C

ontin

ued

approaches and 13 discrete feasible study designs devel-oped:1. Survey approach (one prevalence survey; two inter-

rupted time series; five cross- sectional studies; one household diary study).

2. Qualitative approach (seven key informant interviews; three focus groups).

3. Implementation research approach (four prospective mixed methods cohort studies; one demonstration project).

4. Comparative effectiveness research (three non- randomised controlled trials (non- RCTs); two RCTs).

5. Economic approach (four cost- effectiveness studies; four costing studies).

6. Surveillance (one sentinel surveillance study).7. Standard patient study (1).

We identified selection bias and detection bias as the primary methodological challenges across mixed methods, quantitative and qualitative studies. Detection bias was driven by concerns around self- reported data that may be prone to social desirability bias when partici-pants provide answers they consider the assessors expect to hear. Selection bias in non- randomised studies limits generalisability and occurs when research participants who agree to participate differ qualitatively from those who do not agree to participate. Non- participation may be highest in those vulnerable populations who fear discrim-ination or have privacy and confidentiality concerns. This would then limit the utility of the results obtained from those who do participate, including in relation to GER.

The most frequent GER standards and principles considered relevant were availability and accessibility, followed by privacy and confidentiality.

dIsCussIonWe developed a structured approach to identification of research gaps and formulation of research questions and study designs during a WHO guidelines development process. Our approach builds on our previous work in this area and is rooted in the current GRADE framework used by WHO, further extending it to incorporate the user perspective and foregrounding GER throughout the process.

To the best of our knowledge, our approach is unique in providing practical systematised steps to research ques-tion formulation during WHO guideline development and elaboration of the design of future studies. Prior work in this area has been sparse and focused on identifi-cation and characterisation of research gaps arising from systematic reviews (not guidelines) and has not described the type of research that is required.20 Given that the aim of identification of research gaps is to reduce waste and increase research value,21 collation and possible registra-tion of research gaps on a publicly accessible platform should be encouraged to realign future studies with the existing body of evidence. This would increase access to suggested methodologies and processes to advance

on July 28, 2020 by guest. Protected by copyright.

http://gh.bmj.com

/B

MJ G

lob Health: first published as 10.1136/bm

jgh-2019-002128 on 30 March 2020. D

ownloaded from

Page 12

12 Siegfried N, et al. BMJ Global Health 2020;5:e002128. doi:10.1136/bmjgh-2019-002128

BMJ Global Health

Tab

le 4

R

esea

rch

gap

s fo

r ov

er- t

he- c

ount

er (O

TC) o

ral c

ontr

acep

tion

Clin

ical

, pro

gra

mm

atic

, val

ues-

bas

ed

or

hum

an r

ight

s q

uest

ion

Cur

rent

WH

O

reco

mm

end

atio

nS

tren

gth

of

reco

mm

end

atio

n

So

urce

or

qua

lity

of

evid

ence

Cri

tica

l hum

an

rig

hts

and

eq

uity

co

nsid

erat

ions

Idea

l stu

dy

des

ign(

s)

Feas

ibili

ty

and

pra

ctic

al

cons

trai

nts

Alt

erna

tive

stu

dy

des

ign

Met

hod

olo

gic

al

issu

es a

risi

ng in

the

al

tern

ativ

e st

udy

des

ign

Ben

efits

ve

rsus

har

ms

PIC

O: s

houl

d O

TC

cont

race

ptio

n b

e m

ade

avai

lab

le w

ithou

t a

pre

scrip

tion?

OTC

ora

l co

ntra

cep

tion

shou

ld

be

mad

e av

aila

ble

w

ithou

t a

pre

scrip

tion.

Str

ong.

Very

low

.P

rivac

y an

d

confi

den

tialit

y;

acce

ssib

ility

.

RC

T co

nduc

ted

in

LMIC

with

incl

usio

n of

out

com

es

to m

easu

re

acce

ssib

ility

, priv

acy

and

con

fiden

tialit

y.

As

a st

rong

WH

O

reco

mm

end

atio

n ex

ists

, it

wou

ld b

e un

ethi

cal t

o re

pea

t an

RC

T of

the

ef

fect

iven

ess

of t

he

inte

rven

tion.

Rep

eate

d c

ross

- sec

tiona

l su

rvey

s (c

an b

e ta

rget

ed

to s

pec

ific

sub

grou

ps

or

of w

omen

in t

he g

ener

al

pop

ulat

ion

incl

udin

g ad

oles

cent

girl

s an

d

youn

g w

omen

) att

end

ing

pha

rmac

ies

to e

valu

ate

priv

acy

and

con

fiden

tialit

y an

d c

hoic

e to

ob

tain

OTC

co

ntra

cep

tion

over

tim

e.

Sel

ectio

n b

ias:

w

omen

who

p

artic

ipat

e in

the

su

rvey

may

hav

e fe

wer

con

cern

s re

gard

ing

priv

acy

and

con

fiden

tialit

y th

an t

hose

who

w

ould

ref

use

to

par

ticip

ate,

lim

iting

ge

nera

lisab

ility

.

Are

the

re q

ualit

y d

iffer

ence

s b

etw

een

OTC

ora

l con

trac

eptio

n m

edic

atio

n an

d

pre

scrib

ed o

ral

cont

race

ptio

n m

edic

atio

n?

Not

ap

plic

able

.

G

DG

.

Not

ap

plic

able

.Th

is is

a q

ualit

y as

sess

men

t co

ncer

n an

d r

equi

res

lab

orat

ory

anal

ysis

and

is o

utsi

de

the

scop

e of

thi

s ar

ticle

. WH

O g

uid

elin

es fo

r na

tiona

l reg

ulat

ory

auth

oriti

es e

xist

to

guid

e na

tiona

l re

gula

tory

con

trol

to

ensu

re s

ubst

and

ard

pro

duc

ts a

re n

ot m

ade

avai

lab

le a

nd t

hat

dos

ages

are

co

rrec

t (s

ee h

ttp

s://

ww

w.w

ho.in

t/m

edic

ines

/are

as/

qua

lity_

safe

ty/q

ualit

y_as

sura

nce/

Goo

dR

egul

ator

y_P

ract

ices

Pub

licC

onsu

lt.p

df).

Wha

t ar

e th

e op

timal

w

ays

to p

rovi

de

advi

ce o

n sw

itchi

ng

to d

iffer

ent

oral

co

ntra

cep

tion

or u

sing

ot

her

cont

race

ptiv

e op

tions

(eg,

via

tex

t m

essa

ging

, in

per

son

in p

harm

acy)

?

Not

ap

plic

able

.

G

DG

.A

cces

sib

ility

and

in

form

ed d

ecis

ion

mak

ing.

Com

par

ativ

e ef

fect

iven

ess

rese

arch

to

com

par

e d

iffer

ent

typ

es o

f m

etho

ds

to p

rovi

de

advi

ce.

A c

ontr

olle

d t

rial

that

may

or

may

no

t b

e ra

ndom

ised

is

feas

ible

. Id

entifi

catio

n or

d

evel

opm

ent

of

mod

es o

f hea

lth

liter

acy

will

be

req

uire

d p

rior

to t

estin

g in

a

cont

rolle

d m

anne

r. In

LM

IC, p

harm

acy

tech

nici

ans

or c

ashi

ers

may

pro

vid

e ad

vice

inst

ead

of

pha

rmac

ists

th

at r

equi

res

cons

ider

atio

n w

hen

pla

nnin

g re

al- w

orld

st

udie

s.

Fid

elity

bia

s: d

eliv

ery

of h

ealth

ad

vice

is

dep

end

ent

on m

ode

of a

dvi

ce. I

f giv

en

by

pha

rmac

ists

, co

mm

unity

he

alth

wor

kers

or

tech

nici

ans,

the

n m

onito

ring

the

fidel

ity

of t

he in

terv

entio

n w

ill

ensu

re t

he r

igou

r of

th

e re

sults

. Out

com

es

mus

t in

clud

e a

mea

sure

men

t of

ac

cess

ibili

ty s

uch

as r

eque

sts

for

and

up

take

of a

dvi

ce.

Con

tinue

d

on July 28, 2020 by guest. Protected by copyright.

http://gh.bmj.com

/B

MJ G

lob Health: first published as 10.1136/bm

jgh-2019-002128 on 30 March 2020. D

ownloaded from

Page 13

Siegfried N, et al. BMJ Global Health 2020;5:e002128. doi:10.1136/bmjgh-2019-002128 13

BMJ Global Health

Clin

ical

, pro

gra

mm

atic

, val

ues-

bas

ed

or

hum

an r

ight

s q

uest

ion

Cur

rent

WH

O

reco

mm

end

atio

nS

tren

gth

of

reco

mm

end

atio

n

So

urce

or

qua

lity

of

evid

ence

Cri

tica

l hum

an

rig

hts

and

eq

uity

co

nsid

erat

ions

Idea

l stu

dy

des

ign(

s)

Feas

ibili

ty

and

pra

ctic

al

cons

trai

nts

Alt

erna

tive

stu

dy

des

ign

Met

hod

olo

gic

al

issu

es a

risi

ng in

the

al

tern

ativ

e st

udy

des

ign

Wha

t ad

vers

e ev

ents

ar

ise

from

pro

vid

ing

oral

con

trac

eptio

n O

TC?

Not

ap

plic

able

.

G

DG

.A

ccep

tab

ility

, in

form

ed d

ecis

ion

mak

ing

and

non

- d

iscr

imin

atio

n.

Con

veni

ence

sa

mp

le a

nd c

ross

- se

ctio

nal s

urve

y of

w

omen

ob

tain

ing

OTC

con

trac

eptio

n at

pha

rmac

ies

to

eluc

idat

e if

they

hav

e ex

per

ienc

ed a

dve

rse

even

ts r

elat

ed t

o th

e O

TC n

atur

e of

the

co

ntra

cep

tion.

This

is a

feas

ible

st

udy

des

ign.

N

ot a

pp

licab

le.

Sel

ectio

n b

ias:

suc

h a

sam

ple

wou

ld o

nly

incl

ude

wom

en w

ho

are

will

ing

to p

rovi

de

info

rmat

ion

and

w

ho fe

el s

uffic

ient

ly

confi

den

t to

voi

ce

thei

r co

ncer

ns;

wom

en w

ho h

ave

alre

ady

exp

erie

nced

st

igm

a or

who

fear

it

are

less

like

ly

to r

etur

n to

the

p

harm

acy

and

so

are

less

like

ly t

o b

e re

pre

sent

ed in

the

sa

mp

le.

Valu

es a

nd

pre

fere

nces

Wha

t ar

e th

e va

lues

an

d p

refe

renc

es o

f w

omen

livi

ng in

low

- in

com

e an

d m

idd

le-

inco

me

coun

trie

s re

late

d t

o O

TC o

ral

cont

race

ptio

n?

Not

ap

plic

able

.

G

DG

and

su

rvey

.A

cces

sib

ility

, p

rivac

y an

d

confi

den

tialit

y, a

nd

non-

dis

crim

inat

ion.

Qua

litat

ive

stud

y us

ing

key

info

rman

t in

terv

iew

s w

ith

wom

en s

eeki

ng

and

/or

usin

g co

ntra

cep

tion

and

w

ho li

ve in

LM

IC

rega

rdin

g th

eir

valu

es

and

pre

fere

nces

to

elu

cid

ate

thei

r ex

per

ienc

es a

nd

obse

rvat

ions

. A

que

stio

n re

gard

ing

rece

ipt

of a

pp

rop

riate

sc

reen

ing

for

cont

rain

dic

atio

ns

will

be

esse

ntia

l to

incl

ude.

This

is a

feas

ible

st

udy.

Not

ap

plic

able

.S

elec

tion

bia

s: a

s w

ith a

ll q

ualit

ativ

e st

udie

s, p

artic

ipan

ts

are

pur

pos

ivel

y se

lect

ed. H

owev

er,

this

is a

n ad

vant

age

in t

his

stud

y d

esig

n as

tho

se w

omen

who

ha

ve c

onsi

der

ed

this

issu

e w

ill h

ave

rich

exp

erie

nces

an

d o

bse

rvat

ions

to

shar

e.

Tab

le 4

C

ontin

ued

Con

tinue

d

on July 28, 2020 by guest. Protected by copyright.

http://gh.bmj.com

/B

MJ G

lob Health: first published as 10.1136/bm

jgh-2019-002128 on 30 March 2020. D

ownloaded from

Page 14

14 Siegfried N, et al. BMJ Global Health 2020;5:e002128. doi:10.1136/bmjgh-2019-002128

BMJ Global Health

Clin

ical

, pro

gra

mm

atic

, val

ues-

bas

ed

or

hum

an r

ight

s q

uest

ion

Cur

rent

WH

O

reco

mm

end

atio

nS

tren

gth

of

reco

mm

end

atio

n

So

urce

or

qua

lity

of

evid

ence

Cri

tica

l hum

an

rig

hts

and

eq

uity

co

nsid

erat

ions

Idea

l stu

dy

des

ign(

s)

Feas

ibili

ty

and

pra

ctic

al

cons

trai

nts

Alt

erna

tive

stu

dy

des

ign

Met

hod

olo

gic

al

issu

es a

risi

ng in

the

al

tern

ativ

e st

udy

des

ign

Hum

an r

ight

s an

d e

qui

tyW

ill w

omen

of a

ll ag

es

be

able

to

acce

ss O

TC

oral

con

trac

eptio

n?

Wha

t b

arrie

rs r

emai

n in

th

e he

alth

care

sec

tor?

Not

ap

plic

able

.

G

DG

.A

vaila

bili

ty,

acce

ssib

ility

and

no

n- d

iscr

imin

atio

n.

Una

nnou

nced

st

and

ard

pat

ient

(s

imul

atio

n) s

tud

y w

here

wom

en o

f d

iffer

ent

ages

pos

e as

use

rs w

ishi

ng

to p

rocu

re O

TC

cont

race

ptio

n in

d

iffer

ent

sett

ings

, re

gion

s an

d

coun

trie

s.

This

is a

feas

ible

st

udy

des

ign.

Whi

le t

his

des

ign

is

feas

ible

, cur

rent

ob

stac

les

to c

ond

uctin

g st

and

ard

p

atie

nt r

esea

rch

incl

ude

auto

mat

ed e

lect

roni

c m

edic

al r

ecor

ds

and

cr

eatio

n of

bel

ieva

ble

id

entit

ies.

How

ever

, thi

s sh

ould

be

less

rel

evan

t fo

r O

TC s

tud

ies

than

med

ical

p

ract

ice

stud

ies

whe

re

sim

ulat

ion

is w

idel

y us

ed.

Mea

sure

men

t er

ror:

th

e p

rese

ntat

ion

and

res

pon

ses

of

the

'sta

ndar

d u

ser'

m

ay d

iffer

bet

wee

n sc

enar

ios

and

se

ttin

gs le

adin

g to

ov

eres

timat

ing

or

und

eres

timat

ing

the

acce

ssib

ility

of

the

OTC

co

ntra

cep

tion.

Thi

s ca

n b

e ad

dre

ssed

by

ensu

ring

adeq

uate

tr

aini

ng fo

r 's

tand

ard

us

ers'

and

dev

elop

ing

tem

pla

tes

for

diff

eren

t re

spon

ses.

Acc

epta

bili

tyW

hat

do

heal

thca

re

pro

vid

ers

know

, th

ink

and

feel

ab

out

pro

visi

on o

f OTC

or

al c

ontr

acep

tion,

es

pec

ially

in lo

w-

inco

me

and

mid

dle

- in

com

e se

ttin

gs?

Not

ap

plic

able

.

S

urve

y.A

cces

sib

ility

, p

rivac

y, t

he r

ight

to

seek

, rec

eive

and

im

par

t in

form

atio

n.

Onl

ine

glob

al c

ross

- se

ctio

nal s

urve

y of

he

alth

care

pro

vid

ers

with

del

iver

y th

roug

h re

leva

nt

orga

nisa

tions

.

In m

any

sett

ings

of

dep

rivat

ion,

and

es

pec

ially

in L

MIC

, on

line

acce

ssib

ility

is

lim

ited

by

cost

s, in

adeq

uate

te

chno

logy

and

lack

of

kno

wle

dge

on

how

to

use

tool

s.

Focu

s gr

oup

s in

p

urp

osiv

ely

sam

ple

d

heal

thca

re p

rovi

der

p

opul

atio

ns a

nd r

egio

ns.

Focu

s gr

oup

s ar

e m

ost

suita

ble

to

driv

e co

nver

satio

ns a

nd

elic

it re

spon

ses

amon

g p

artic

ipan

ts w

ith s

hare

d

lived

exp

erie

nces

and

is

less

con

fron

tatio

nal a

s th

e ai

m is

not

to

gath

er in

- d

epth

per

sona

l exp

erie

nces

b

ut a

sses

s cu

rren

t cu

ltura

l an

d s

ocia

l nor

ms,

pra

ctic

es

and

con

cern

s w

ith r

esp

ect

to O

TC c

ontr

acep

tion.

S

pec

ific

que

stio

ns c

an b

e d

irect

ed r

egar

din

g at

titud

es

tow

ard

s vu

lner

able

p

opul

atio

ns, f

or e

xam

ple

, tr

ans-

men

and

bis

exua

l w

omen

.

Soc

ial a

ccep

tab

ility

b

ias:

giv

en t

hat

exp

erie

nces

are

sh

ared

in a

gro

up,

ther

e m

ay b

e a

relu

ctan

ce t

o d

iscl

ose

stig

mat

isin

g b

ehav

iour

s; h

owev

er,

dev

elop

men

t of

se

mis

truc

ture

d

inte

rvie

w q

uest

ions

in

con

sulta

tion

with

lo

cal h

ealth

care

p

rovi

der

s to

bes

t re

flect

loca

l bel

iefs

fo

r in

terr

ogat

ion

can

red

uce

this

bia

s. In

ad

diti

on, a

ski

lled

fo

cus

grou

p le

ader

ca

n cr

eate

a s

afe

spac

e fo

r d

iscl

osur

e.

Tab

le 4

C

ontin

ued

Con

tinue

d

on July 28, 2020 by guest. Protected by copyright.

http://gh.bmj.com

/B

MJ G

lob Health: first published as 10.1136/bm

jgh-2019-002128 on 30 March 2020. D

ownloaded from