Cultural adaptation of the person-centered maternity care scale at governmental health facilities in Cambodia

Background In Cambodia, the importance of valuing womens childbirth experiences in improving quality of care has been understudied. This is largely because of absence of reliable Khmer tools for measuring womens intrapartum care experiences. Generally, cross-cultural development of those tools often involves translation from a source language into a target language. Yet, few earlier studies considered Cambodian cultural context. Thus, we developed the Cambodian version of the Person-Centered Maternity Care (PCMC) scale, by culturally adapting its original to Cambodian context for ensuring cultural equivalence and content validity. Methods Three rounds of cognitive interviewing with 20 early postpartum women were conducted at two governmental health facilities in Cambodia. Cognitive interviewing was composed of structured questionnaire pretesting and qualitative probing. The issues identified in the process of transcribing and translating audio-recorded cognitive interviews were iteratively discussed among study team members, and further analyzed. Results A total of 14 issues related to cultural adaptations were identified in the 31 translated questions for the Cambodian version of the PCMC scale. Our study identified three key findings: (i) discrepancies between the WHO recommendations on intrapartum care and Cambodian field realities; (ii) discrepancies in recognition on PCMC between national experts and local women; and (iii) challenges in correctly collecting and interpreting less-educated womens views on intrapartum care. Conclusion Not only womens verbal data but also their non-verbal data and cultural contexts should be comprehensively counted, when reflecting Cambodian womens intrapartum practice realities in the translated version. This is the first study that attempted to develop the tool for measuring Cambodian womens experiences during childbirth, by addressing cross-cultural issues.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022.   is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022. care guidelines for a positive childbirth experience to enhance women-centered outcomes [2]. 69 Earlier studies suggested that the person-centered dimensions determine women's decisions to 70 seek health care [3][4][5][6][7][8][9]; their intentions to select the same facility for future deliveries [10, 11]; 71 their demands for more friendly care [3, 12, 13]; their satisfaction with quality of care [4,11,72 14]; and long-term impacts on their future reproduction [15,16]. 73 Cambodia is a lower-middle-income country in South East Asia that has undergone 74 significant hardship such as colonialization, civil wars, and genocide during the Khmer Rouge 75 regime. Note that Cambodian health system was entirely damaged and even destroyed (incl. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint scale, i.e. "0 = No, never", "1 = Yes, a few times", "2 = Yes, most of the time", "3 = Yes, all 100 the time". The full scale and subscales showed good internal consistency reliability in Kenya, 101 India, and Ghana, with a Cronbach's α value of over 0.8 for the full scale and ranging between 102 0.61 and 0.75 for the subscales [24].

103
When developing a cross-cultural tool, multi-dimensional factors need to be considered 104 not only in translation but in cultural interpretation, values, and attitudes [25]. Moreover, cross-105 cultural researches generally require strategies for adequately addressing the fact that a concept 106 in a certain setting may not be identical or comparable to that in another setting and that a tool 107 appropriate in one context could be inadequate in another context [26]. A questionnaire 108 translated into another language must meet the following requirements: (i) be valid, reliable, 109 and cost-effective; (ii) exhibit appropriate levels of semantic and conceptual equivalence [27].

110
In developing a cross-cultural tool involving translation from an original language to another . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint domicile, diverse economic levels, different religions, and both low and high-risk pregnancies).

146
The health center serves 10,000-20,000 population in its catchment area, and provides antenatal   is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint translated into Khmer, parallelly by two independent Khmer-English bilingual translators 159 whose mother tongue is Khmer. The two versions were then synthesized by a Cambodian Khmer monolinguals. Ambiguous and unmatched texts were thoroughly discussed and 164 addressed among the study team members including the trilingual principal investigator whose 165 mother tongue is Japanese, to generate the Kh-PCMC scale version 2. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022.  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint   (Table 3). is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022. ; https://doi.org/10.1101/2022.03.09.22272127 doi: medRxiv preprint  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022. ; https://doi.org/10.1101/2022.03.09.22272127 doi: medRxiv preprint Table 3 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint found that six of those 12 issues at comprehension stage required cultural adaptation. S1 Table   245 presents the process of translation into Khmer and the reasons for revisions.
. CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint  Did the medical staffs call you by your name?
Did the medical staffs call you appropriately? reason, midwives did not introduce themselves to women, when they visited the health center.

258
The results of qualitative probes however revealed that nine of 20 respondents wanted to know 259 which midwife would be responsible for their childbirths so they could call them when there is 260 a problem. Thus, we re-translated "welcome" back to "introduce themselves" and added an 261 explanation. The response errors were addressed by the additional explanations.

262
Similarly, Item #3 "being called by name" was replaced by "called appropriately" in the 263 translated based on the experts' suggestions. But it was found that 11 of 20 respondents were . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Item #10 "delivery position choice" did not make sense among respondents because it is 297 common to give birth in the supine position in Cambodia. Thus, "choice" was replaced by 298 "favorite free position" that is more understandable to Cambodian women. It was found in 299 qualitative probes that five of 20 women preferred free movements during labor to reduce labor 300 pain and contribute to smoother delivery. One woman at the urban hospital reported that free 301 delivery position was explained to her during delivery, by using its health education material 302 (posters). As a result, she could freely move during labor. For this reason, "favorite free position" 303 was employed as a replacement for "choice".

304
When responding to Item #19 "when you needed help from medical staff", three of 20 305 respondents at the urban hospital misinterpreted the item as need for help from their family 306 members. We reminded them that the item implies help not from family members but from 307 medical staff, when this misinterpretation occurred. By doing so, mothers' cognitive response 308 errors were adequately addressed. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022. ; https://doi.org/10.1101/2022.03.09.22272127 doi: medRxiv preprint 25 312 know whether their health information was kept confidential. While, five of 20 respondents 313 mentioned that no information was kept confidential (incl. household poverty specified in 314 poverty ID card and HIV status). All the 20 respondents stated that they wanted to share their 315 personal health information with their family members. In view of these statements, another 316 response category "don't know if it was kept secret" was added as a response category.

318
Women's cognitive issues 319 The respondents appeared to have difficulties in selecting one out of multiple answer 320 options in frequency-scale-based items. Although the interviewers read aloud each response 321 option, the respondents often stated just "Yes" without selecting any specific option given. confidentiality. Yet, it found later that she did not even know whether the information was kept 327 confidential. Thus, supplementary explanations were often necessary for some questions. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Moreover, it was found that many respondents found it difficult to correctly understand 329 the questions on abstract concepts, for instance when asked the question "what respect means?". 330 Many women mentioned that it was good enough that medical staff were ready to see them on 331 time.

334
Mean number of minutes spent on a cognitive interviewing was 65 minutes (range: 40-335 90 minutes). Overall, the length of interview time was acceptable to respondents, although some 336 families complained that it was too long. All the 20 respondents preferred to have a face-to-337 face interview, though it was more time-consuming than self-administration of the 338 questionnaire. Some stated "…Because it is difficult for me to read through the questionnaire…" 339 and "…Because I have never experienced in self-administering any questionnaire before…". 340 Those with no formal education took longer time than those with formal education. An example 341 is that less educated women were neither experienced in nor familiar with selecting one of the 342 four response options for the questions. One woman without formal education repeatedly stated 343 "Thank you" without responding to the questions, because she was so delighted and honored to . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022.  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint This result is in line with an anthropological study that Cambodian people believe the human 380 being inseparably consists of physical body and name [47]. In Cambodian context, the body is 381 a human being's container, while its name is given to the body. This inseparable unit identifies 382 a "self" that distinguishes from others. One of the important foundations for correctly 383 interpreting Cambodian contexts is Buddhism sense of self such as "reincarnation" and "karma".

384
It is believed in Cambodia that every woman has eight reincarnation stages for eight rites of 385 passage during lifetime from birth to death [47]. Childbirth is one of eight rites of passage for 386 a woman. It is also believed that a woman's body at the current reincarnation stage will be 387 reborn through childbirth and her name should be religiously changed when entering the next is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022. ; https://doi.org/10.1101/2022.03.09.22272127 doi: medRxiv preprint 31 409 reported male-dominant gender norm significantly influences decision-making process [41]. In 410 fact, we encountered the situation where husbands refused to let their wives participate in this 411 study in the consent process. This is largely because women's inability to participate in decision 412 making is socially institutionalized. This is one of the significant discrepancies between 413 globally emphasized importance of women's self-determination and local gender norm. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022. ; https://doi.org/10.1101/2022.03.09.22272127 doi: medRxiv preprint 32 425 evidence-based local implementation needs to be accelerated in each existing health system. 426 Fourth, regarding Item #19 "when you needed help from medical staff", some women 427 were confused with the question about need for help by misinterpreting help from family 428 members. This can be explained, to a large extent, by task sharing between doctors, midwives, 429 nurses, and women's family members that were derived from shortage of health workforce.  Fifth, it was found that Item #7 "record confidentiality" was often not a concern. For 436 instance, even data on household's economic level and individual's HIV status were shared and is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint the PCMC scale provides an opportunity to further explore respondents' initial response.

493
In addition, it was challenging to elicit rich qualitative data from less-educate women. 494 We failed to elicit deeper meanings of "respect" despite repeated qualitative probes. This is 495 probably because many respondents had difficulties in understanding abstract concept such as  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint To fully describe Cambodian women's realities, it is important to triangulate non-verbal 508 data and description of the cultural context rather than relying on only verbal data from women. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint   is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint identifying methodological issues in developing a cross-cultural tool. This study also highlights 539 the importance of cross-cultural translation and adaptation, before formally using the 540 questionnaire in a survey. Translation and adaptation processes require not only technical and 541 linguistic skills and experiences, but also considerable time necessary for ensuring cross-542 cultural equivalence. A further study needs to be conducted, to assess broader landscapes and 543 greater perspectives. Further, the use of a variety of data including anthropological approaches 544 is key to fully capturing Cambodian women's reality.  Ganchimeg. We are grateful to all the people who contributed to this study at various stages 551 in developing the Kh-PCMC scale, including translators, Cambodian experts, a field 552 coordinator, field assistants, and study participants in Cambodia. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022. ; https://doi.org/10.1101/2022.03.09.22272127 doi: medRxiv preprint (37) Van de Vijver, Fons JR, Leung K. Methods and data analysis for cross-cultural research.: . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 10, 2022. ; https://doi.org/10.1101/2022.03.09.22272127 doi: medRxiv preprint