Socio-anthropological methods to study the feasibility and acceptability of the minimally invasive autopsy from the perspective of local communities: lessons learnt from a large multi-centre study

Figures & data

Table 1. Site characteristics and description.

Country	Location	Population	Urban/Rural	Main religion	Main income activity	Ethnolinguistic groups	Health facilities	Post-mortem procedures
GABON	Lambaréné	70,000	Semi-urban	Christian	Farming, Fishing	Omyene, Fang, Guischira, Tsogo	District hospital (2), Healthcare centre (2), Community Clinic (10)	Verbal autopsies
KENYA	Siaya county, Kisumu	227,400	Rural	Christian	Farming, Fishing	Luo	County Referral Hospital (1), Healthcare centre (30)	Research-related CDAs^a
MALI	Bamako	1,810,366	Urban	Muslim	Trade, Industry, Administration, Agriculture, Farming, Crafts	Bamanans, Peuhl, Sonrais, Malinké, Soninké, Sénoufos	Tertiary hospital (1), Healthcare centre (3)	Verbal autopsies, Non-routine CDAs^b
MOZAMBIQUE	Manhiça	160,000	Rural	Christian	Farming, Industry, Informal trading	Shangaan	District hospital (1), Rural hospital (1), Healthcare centre (12)	Verbal autopsies
	Maputo	2,000,000	Urban	Christian	Trade, Industry, Administration	Shangaan	Quaternary hospital (1)	Routine CDAs, Research-related MIAs
PAKISTAN	Karachi	14,910,352	Urban	Muslim	Trade, Industry, Administration	Urdu, Sindhi, Punjabi, Pashto, Seraiki	University hospital (7), Tertiary hospital (3), numerous private health centres	Non-routine CDAs^b

^aResearch Study on tuberculosis

^bForensic CDAs

Table 2. Study participants per site.

		Final participants per country
Group	Expected size per country	GA	KE	ML	MZ	PK	Final participants all countries
Key informants	30	26	88	23	31	29	197
Healthcare providers	30	29	53	37	33	31	183
Next of kin who had suffered the death of a relative	30	29	54	31	46	34	194
In the preceding 0–24 hours	10	4	4	2	9	5	24
In the preceding 1–7 days		9	17	14	25	12	77
In the preceding 30–40 days	20	16	33	15	12	17	93
TOTAL	90	84	195	91	110	94	574

GA: Gabon; KE: Kenya; ML: Mali; MZ: Mozambique; PK: Pakistan

Table 3. Number of events per site.

Site	IDI	SSI	FGD	OBS	Case study	Total N of events
Lambaréné, GA	42	42	4	11	0	99
Siaya county, Kisumu, KE	63	66	6	5	0	140
Bamako, ML	38	53	0	23	0	114
Manhiça & Maputo, MZ	39	67	0	45	4	155
Karachi, PK	46	48	0	20	0	114
Total	228	276	10	104	4	622

Panel 1. Useful findings and lessons learnt during the implementation of the F&A multi-centre study.

Master Protocol and general socio-anthropological methods

If more than one site, it is recommended to use a systematic approach (Master protocol), agreed by all sites’ team, and flexible enough to include all local idiosyncrasies The mix of different qualitative data collection methods (IDI, SSI, FGD, observations, case studies) provides complementary information, and is therefore highly recommended Triangulation of the data provides validity and increases the general discernment of the major emerging themes from the study Data should be processed, transcribed, translated and coded locally, in order not to lose meanings and views

Preparation and implementation of activities

For a study involving the use of any given post-mortem procedures, particularly in places where such methods have seldom been used, a community entry and engagement strategy is essential Preparatory meetings with community and religious leaders and CABs may provide useful recommendations on how to conduct data collection and how to interact and behave when approaching interviewees/observations Flexibility for adapting or changing study procedures according to these recommendations is important

Team training and preparedness

The very sensitive nature of the procedures and circumstances around death implies an even higher need for continuous training and preparedness of staff Psychological support may be required to diminish the consequences in study staff of taking part in post-mortem procedures and burial ceremonies; and communication with grieving families The risk of observation biases derived from such psychological impact in terms of the interpretation of data collected needs to be minimized

Choice of Informants/Stakeholders

Targeting different kinds of informants bring complementary perspectives Building rapport with healthcare providers may be challenging, and planning a specific mobilization strategy may be necessary

Interview planning

An initial prudent approach to assess such a sensitive phenomenon as death is recommended, in order to minimize negative responses For next of kin, teams need effective systems to rapidly identify participants

Interviews with next of kin of deceased individuals

Approaching family members is challenging but feasible if done with respect Phasing in from 30–40 days interviews to 1–7 days was useful to the team Counter-intuitively, interviews with family members in the first 24 h after death were easier to conduct than those at later stages

Observations

Observations provided extremely valuable information, and as a result were increased in number and nature beyond what was originally planned For observations, a different set of skills is required from the study team Start with more ‘public’ observations (cemeteries, etc.) and only move to more private ones (washing of bodies, etc.) once the study staff feels comfortable A right mix of participant and direct non-participant observations is useful Biosafety measures that are culturally appropriate must be offered to field ethnographers to ensure their adequate protection against communicable diseases during peri-mortem procedures or ceremonies and cemetery events

Feedback to the community

A good relationship with the community implies being available for responding any doubts, and promptly providing feedback to community and families

Initial assumptions proven wrong

Approaching the families was less difficult than expected Refusals to participate in the study were lower than expected It was easier to speak with the next of kin of a person deceased in the preceding hours than to that of someone having died in the preceding month The term ‘Autopsy’ and its implications was better understood at the community level due to already existing traditional procedures that implied the handling of corpses (traditional autopsies and/or traditional caesarean sections) The term ‘Autopsy’ was not necessarily perceived to have negative connotations