Measuring Prevalence of Disrespect and Abuse during Childbirth in Tanzania and Ethiopia: The Hansen Project Hannah Ratcliffe June 24, 2014
Feb 24, 2016
Measuring Prevalence of Disrespect and Abuse during Childbirth in Tanzania and Ethiopia:
The Hansen Project
Hannah Ratcliffe June 24, 2014
Partnerships
Study Areas
Dar es Salaam, Tanzania Amhara & SNNP Regions, Ethiopia
Photo credit: Brenda D’mello, CCBRT Photo credit: Kelly Ramundo/USAID
Study Design
• Baseline study – TZ: April – August, 2013; ET: July – September, 2013– What are types and prevalence of disrespect and abuse
among women delivering in public health facilities?– What are factors associated with disrespect and abuse?
• Intervention Phase (ongoing)– Which locally acceptable, effective, and sustainable
intervention(s) can reduce disrespect and abuse during childbirth?
• Evaluation
Research Methods
Quantitative Method Tanzania EthiopiaExit InterviewsClients interviewed at time of discharge from facility—approximately 3-6 hours post-delivery N=2000 N=204
Direct observations of client-provider interactions Clients observed throughout their time at the facility, from registration to 2 hours post-delivery N=208 N=193
Community Follow-Up InterviewsMothers interviewed in their homes approximately 4-6 weeks post-delivery N=70 N/A
Provider InterviewsProviders interviewed and completed self-administered survey on job satisfaction and perceptions of clients N=50 N=30
Seven Categories of Disrespect and Abuse
Category ExamplePhysical abuse Slapping, pinching
Non-consented care Absence of informed consent or patient communication, forced procedures
Non-confidential care Lack of privacy (e.g. laboring in public) and/or confidentiality (e.g. disclosure of patient information)
Non-dignified care Intentional humiliation, rough treatment, scolding, shouting, blaming, negative perceptions of care
Discrimination based on specific patient attributes
Discrimination based on race, ethnicity, age, language, HIV status, economic status, educational level, etc.
Abandonment of careWomen left alone during labor and birth, failure of providers to monitor women and intervene when needed
Detention in facilities Detention of mother and/or baby in facility after delivery, usually due to failure to pay
Bowser and Hill 2010
Operationalizing the 7 Categories
Bowser & Hill Category
Hansen Category Example
Physical abuse Physical Abuse Slapping, pinching
Non-consented care
Non-consented care
Absence of informed consent or patient communication, forced procedures
Non-confidential care
Non-confidential care
Breaches of auditory privacy, eg disclosure of patient information while others can hear
Lack of privacy Breaches in physical privacy, eg laboring in public, insufficient privacy infrastructure (curtains, partitions)
Non-dignified care Non-dignified care Intentional humiliation, rough treatment, scolding, shouting, blaming, negative perceptions of care
Discrimination based on specific patient attributes
N/A
Abandonment of care
Abandonment of care
Women left alone during labor and birth, failure of providers to monitor women and intervene when needed
Detention in facilities
Detention in facilities
Detention of mother and/or baby in facility after delivery, usually due to failure to pay
Calculating Prevalence
• Cross-country and cross-regional comparisons using a variety of methods have yielded several lessons learned:– Context matters– Timing and setting matter– Question phrasing matters
• And several outstanding questions:– Whose opinion counts?– How is reporting influenced by normalization and biases?
Calculating Prevalence—Context Matters
Type of Disrespect and AbuseTanzania
N=2000, n (%)Ethiopia
N=204, n (%)Any form of disrespect or abuse 292 (15) 43 (21)
Physical Abuse 90 (5) 1 (0.5)
Non-Consented Care 5 (0.3) 36 (18)
Non-Confidential Care 34 (2) 28 (14)
Lack of Privacy 37 (2) 31 (15)
Non-Dignified Care 125 (6) 2 (1)
Abandonment 154 (8) 5 (3)
Detention 4 (0.2) 0 (0)
Differences between countries at exit interview:
Calculating Prevalence—Context Matters
Type of Disrespect and Abuse
Ethiopia Overall
N=204, n (%)
Amhara Region
N=98, n (%)
SNNPRegion
N=106, n (%)Any form of disrespect or abuse 43 (21) 38 (39) 5 (5)
Physical Abuse 1 (0.5) 1 (1) 0 (0)
Non-Consented Care 36 (18) 33 (34) 3 (3)
Non-Confidential Care 28 (14) 27 (28) 1 (0.9)
Lack of Privacy 31 (15) 29 (30) 2 (2)
Non-Dignified Care 2 (1) 2 (2) 0 (0)
Abandonment 5 (3) 5 (5) 0 (0)
Detention 0 (0) 0 (0) 0 (0)
Differences by region within countries:
Calculating Prevalence—Timing & Setting Matter
Differences between reporting at Exit and CFU interviews:
Type of Disrespect and Abuse
TanzaniaExit InterviewN=2000, n (%)
TanzaniaCFU
N=70, n (%)
Any form of disrespect or abuse 292 (15) 54 (77)
Physical Abuse 90 (5) 36 (51)
Non-Consented Care 5 (0.3) 4 (6)
Non-Confidential Care 34 (2) 36 (51)
Lack of Privacy 37 (2) 36 (51)
Non-Dignified Care 125 (6) 37 (53)
Abandonment 154 (8) 35 (50)
Detention 4 (0.2) 1 (1)
Calculating Prevalence—Phrasing Matters
Question
TanzaniaExit InterviewN=2000, n (%)
Experienced D&A At any point during your stay for this delivery, were you ____/ did you experience anything that ___?
292 (15)
Perceived D&A On a scale of 1-5, were you treated in a way that made you feel humiliated or disrespected?
344 (17)
Calculating Prevalence: Whose opinion matters?
Type of Disrespect and Abuse
TZ ExitN=2000
n (%)
TZ Obs.N=70 n (%)
ET ExitN=204
n (%)
ET Obs.N=193
n (%)Non-Confidential Care 34 (2) N/A 28 (14) N/A
Discussed personal issues in earshot of other clients
1 (0.1) --- 3 (2) ---
Mother’s history taking findings shared when others could hear
--- 42 (20) --- 64 (33)
Non-Consented Care 5 (0.3) N/A 36 (18) N/A
Non-consented vaginal examination 5 (0.3) --- 31 (15) ---
Lack of consent for vaginal examination
--- 170 (82) --- 132 (68)
Comparing prevalence from interviews and observations:
Reporting: Normalization and Bias
• “Prevalence” from exit interview/CFU is actually reported prevalence– Reporting is as complicated and personal as the
experiencing of the event• Evidence of possible normalization—comparing
observation and exit interviews• Bias in reporting caused by:– Post-delivery factors: exhaustion, relief, etc.– Fear of retribution
Conclusions and Next Steps
• Will report “prevalence” primarily as experienced D&A from exit interviews and community follow-up
• Working to determine an appropriate way to calculate observed prevalence
• Evaluation in Tanzania will be based on community follow-up– Evaluation design in Ethiopia still pending
Acknowledgements
• John and Katie Hansen Family Foundation
• Tanzania Ministry of Health and Social Welfare
• Ethiopia Federal Ministry of Health