Parent feedback: perception of parents and pediatric residents in Asian culture

Residency programs have used feedback of parents as a part of multisource feedback to improve residents’ skills. However, there is no evidence showing how parents can appropriately be engaged in pediatric resident feedback in our context. This study aimed to investigate: 1) the resident skills assessed by parents in perception of Thai parents and pediatric residents and 2) the parent characteristics making credible feedback in perception of Thai pediatric residents.


Abstract Background
Residency programs have used feedback of parents as a part of multisource feedback to improve residents' skills. However, there is no evidence showing how parents can appropriately be engaged in pediatric resident feedback in our context. This study aimed to investigate: 1) the resident skills assessed by parents in perception of Thai parents and pediatric residents and 2) the parent characteristics making credible feedback in perception of Thai pediatric residents.

Methods
The participating parents were asked to rate their con dent to assess residents' skills according to 17 items of the PARENTS questionnaire (Likert scale [1][2][3][4][5]. To enhance the reasons of parent's con dent, the parents responded to think aloud interviews. We conducted focus groups in order to explore the resident perception on parents' assessable resident skills and credible parent characteristics.

Results
There were 51 parents revealing the mean con dent scores of 17 items ranged from 4.06 to 4.37. The parents felt con dent to assess the residents' skills according to the questionnaire because they directly experienced the residents' performance during admissions. Five resident focus groups were conducted. In resident perception, residents' skills in communication, collaboration, leadership and professionalism were assessable by parents. The characteristics affecting the feedback credibility included parents' prior experience, personal behavior and confrontation to medical situation.

Conclusions
Thai residency programs may use parent feedback for residents' performance improvement as both parents and residents are able to identify the parents' assessable skills. The feedback process should consider the credible parents' characteristics to increase the acceptability and avoid the bias.

Background
In pediatrics, parents or caregivers often have direct interactions with pediatricians because they are always being with their children. They observe how physicians take care of the patients, communicate with them and other professions, and demonstrate professionalism [1]. Parents are willing and able to assess some physicians' skills. Pediatric residents positively react to parent feedback and used it for assessing and improving their performance [2,3]. Parent feedback has been widely studied and used in Western countries whereas there are only inadequate studies in Eastern countries. Perceptions on using Hofstede model has been used to explain cultural differences between countries [12]. Hofstede classi ed cultural differences between countries using ve dimensions: power distance, individualism, uncertainly avoidance, masculinity, and long-term orientation. The Hofstede's cultural dimensions theory describes the effects of a society's culture on values of its member and how these values relate to behavior. Suhoyo, et al. found cultural differences in feedback processes during clerkships comparing between Indonesian and Dutch medical students [13]. They found that students in the country with high power distance and low individualism (Indonesia in this case) preferred to receive feedback from specialists and perceived feedback from specialists and residents as more instructive than feedback from nursing and paramedical staff. In contrast, Dutch students valued the feedback from residents and specialists equally. They perceived the feedback from direct observation was more instructive. Moreover, cultural differences also in uence the perceived learning value of feedback. Indonesian medical students recognized characteristics of four valuable feedback consisting of weaknesses mentioned by provider, comparing performance to a standard, explaining or demonstrating the correct performance, and preparing an action plan with the student [14]. Appraisal of good performance, another valuable characteristic of feedback, was not valued from Indonesian students. Cultural differences seem to in uence the perception in credibility of feedback sources and characteristics of feedback of population in each culture. Thailand is classi ed to have high power distance dimension. Thai people are also more collectivist. This characteristics are opposite to the Western's cultural dimensions. The perception of parent feedback in such high power distance and collectivist culture has never been investigated.
To effectively engage parents to be a source of pediatric residents' MSF, parents on their side need to be con dent to assess physicians' skills while pediatric residents need to trust parents as a credible source. The purpose of this study was to explore the perception of parents on their assessable physicians' skills and of pediatric residents on credibility of parent feedback in our context. This is an important topic to be studied because we will be able to use parents as a source of MSF in the most effective and appropriate way. The results of this study may be applied to other countries with the same cultural dimensions.
In order to respond to the purpose, this study asked the following questions:

Participants
Two groups of purposive participants were invited to the study.
1. Parents or caregivers of children and adolescents (age less than 18 years) with a history of more than two in-patient admissions due to acute or chronic illnesses: These parents are likely to be familiar to the hospital systems and roles of health care team members. These parents should thus be able to recognize the team members such as medical students, residents, faculty, and nurses because they have communicated and worked with them during their child admissions.
Parents were directly asked to participate in the study by the researcher. This study expected to recruit 50 parents or caregivers to participate in the quantitative part (responding to the questionnaire). Parents were also asked for a think-aloud interview after answering the questionnaire. The nal number of parents in this qualitative part would be determined when the data was saturated. Residents were informed about this study by the researcher after a morning educational activity.
Residents, then, were invited via email. The nal number of focus groups was determined when data saturation was reached.

Design
To explore the parent perception, this study used an explanatory sequential mixed methods consisting of two parts. The rst part was a quantitative study. An existing parent feedback questionnaire was used to identify assessable physician skills. The second part was a qualitative study in order to further explore the parent perception on their con dence to providing feedback. To study the resident perception on parents' assessable skills of the residents and the credible parents' characteristics, we used a qualitative method.

Data collection
The quantitative part

Parent characteristics
Characteristic data of enrolled parents were collected. The data included age, relation to the patient (mother, father, or caregiver), number of children, educational background (high school, Bachelor, Master, or PhD degree), career, and economic status. Disease status of the patient (acute or chronic disease) was also recorded.
The PARENTS questionnaire An existing parent feedback questionnaire assessing pediatric residents' skills was selected: Parents' Assessment of Residents Enacting Non-Technical Skills (PARENTS). This questionnaire was developed from parent assessment of residents in a pediatric emergency departments [11]. The PARENTS consists of 17 close-ended and 2 open-ended items with 5-point Likert-type rating scales. This tool has been used to assess residents' skills including communication, comfort in a pediatric setting, adaptability, and collaboration. This tool showed high validity evidence. Although this questionnaire was developed in the emergency department setting, the items assessing resident skills were quite general. Therefore, this questionnaire should also be used in general pediatric setting. The original principle investigator of this questionnaire was asked for permission of use. Although this questionnaire is used for assessing residents' skills, this study used the questionnaire, however, not to judge residents, but to test the parents' con dence in answering the questions in the questionnaire. The questionnaire was translated from English into Thai language. The processes of translation were adapted from the guidelines for translating a questionnaire into a different language [17].
The questionnaire was administrated by the researcher. Parents were asked to read each item of the questionnaire. Then, parents responded how they felt con dent to evaluate the residents' skills according to each item. The answer options of each item was a 5-level Likert scale ranging from "no con dent" (score 1) to full con dent (score 5). The parents replied with the level of their con dence to assess residents' skills but not actually judge the performance of residents. So this rating scale differed from the original scale of the PARENTS instrument.

The qualitative part
The qualitative part consisted of two tasks. The rst task was to explore in-depth data on parents' con dent of assessable skills. The second task was to investigate the perception of pediatric residents on parents' assessable skills and credibility of parents' feedback.

Parent perception
The parents were invited for think-aloud interviews in order to in-depth explore on their con dent of assessable skills. The parents voluntarily participated in the interview part after answering the questionnaire. The researcher used probe questions eliciting detailed information beyond the answers given by the parents. The probe question of each item was "why did you answer this option?". After nishing all 17 items, the researcher asked "which any other skills or abilities of residents do you think you can assess?". The interviewing was audiotaped and transcribed verbatim.

Resident perception
The researcher invited pediatric residents for focus groups in order to get information from different participants' opinion. Each group consisted of 6-8 participants. The members of each group were the same year residents in order to avoid power differences. Almost all involved residents were females (41 females and 4 males) and all rst year residents were females. Therefore, there was no equal male/female ratio possible in each group. The researcher was an interviewer or moderator and a colleague was an observer. The observer helped picking up signi cant non-verbal communication in the group. Although there was a power distance between the residents and interviewer, it could only have minimal effect on the interview process because the discussion issues did not involve the relationship between the residents and moderator. The discussion guide with a list of questions was prepared in advance [16]. The interview guide is shown in Table 1. The expected time for each focus group was about 45 minutes to one hour. The content of interviews and focus groups was audiotape recorded and transcribed verbatim.

Data analysis
Quantitative part Characteristics of the parents were presented by category and frequency. The mean (S.D.) of the con dent scales of each item and mean overall score were calculated. Comparison of the mean overall scores of selected parents' characteristics were analyzed using t-tests.

Qualitative part
Data collection and data analysis occurred concurrently. The new data or information might emerge and was used for formulating new questions or performing member checking [17]. Iterative data collection and analysis was continuously performed until reaching data saturation.
Using content analysis approach, data analysis was performed by coding and categorizing from participants' narratives [18]. Finally, categories formed themes. A second researcher with experience in conducting qualitative research coded and categorized transcripts independently. Then, we met to compare and discuss the coding and categorization. Re exivity was recorded on diary during data collection and analysis.

Ethics
This study was approved by Ethical Committee for Research of Faculty of Medicine Ramathibodi hospital before collecting the data. Each participant was asked for written informed consent.

Parent perception on residents' assessable skills
There were 51 parents participated in this study. The mean age of the parents was 40.7 years. Most of caregivers were mothers. Sixty percent of the parents graduated higher education degrees. All of the admitted patients were diagnosed with chronic diseases and most of them had history of more than 10 admissions. The characteristics of the parents are shown in Table 2. Parents read the PARENTS questionnaire and checked their con dent scores from 1 to 5 for each item. Most parents felt con dent at the level of 4 or 5 for each item ranging from 82.3-96.1% (Table 3). The mean (SD) of the overall scores was 4.18 (0.80). No more than 2% of the parents selected the con dent level 1 or 2 for any of the questions. For questions 4, 10, 13, and 15, more than 10% of the parents answered level 3. For two questions, 3 and 9, more than 40% of the parents answered level 5. There was no association between educational level or number of admissions and the mean scores (Table 4). The mean of the overall scores of parents with high school graduation or less (4.05) was not different with the mean of the overall scores of parents with higher education graduation (4.27). Parents experiencing 10 or more admissions had the mean of the overall scores of 4.17 which was comparable to the mean of the overall scores of parents with less than 10 admission experiences (4.22). Parent think-aloud interview on residents' assessable skills A total of 21 parents were interviewed. These parents were asked to explain how they were or were not con dent to assess residents' skills designated in each item of the questionnaire. Parents' perception and quotes are shown in Table 5. Quotes are translated from Thai, individual parents are indicated anonymously by [Px]. However, some parents were not con dent that the residents understood their needs.
"They (residents) listen to us. If they listen to us, they will make correct diagnosis and provide appropriate treatment." [P5] "….. I am not con dent. I am not sure whether they (residents) understand what I say to them." Residents informed parents about procedures which they were going to do with the patients. Investigations and plan of treatment were also the two most frequent issues which residents discussed with parents and patients. 10 Parents saw residents' exibilities of their thinking and approach when they asked the residents for changing the plans of investigations or treatment.
"My daughter gets venipuncture every morning. She suffers a lot. So I ask the doctor (resident) to collect a bunch of tests at one time ….." [P11] 11 Pediatric residents frequently asked about parents' and patients' well-being and supported them. Parents saw the resident's facial expression and felt their empathy.
Some parents felt that primary doctors of their children showed more concern compared to the residents.
"I can touch that feeling. When we feel sad, we always receive support from them (residents)." [P11] "Yes, they concern ….. but it may be not similar to our relatives or primary doctor of my child" Residents' perception on assessable skills Thirty six residents participated in focus groups. There were 5 focus groups: group 1 (n = 7) and group 2 (n = 8) of pediatric resident year 1, group 3 (n = 8) of year 2 and group 4 (n = 7) and group 5 (n = 6) of year 3. The participating residents identi ed four skills which would be able to be assessed by parents including communication, collaboration, leadership and professionalism.

Communication skills
All residents perceived that parents could well assess their communication skills. Residents described the daily activities they interacted with patients and parents which needed effective communication. They frequently informed medical information including clinical progression, results of investigations, plans of treatment and prognosis. Moreover, the residents spent more times with anxious parents and complicated cases. The residents highlighted components of communication skills which parents were able to assess: (1) abilities to make parents understanding the medical information; (2) active listening skills; and (3) empathy expression.
Residents' abilities to make parents understanding the medical information Residents informed the signi cantly medical information to parents every day. Parents could realize how well they understood the information they received from the residents. Parents' understanding the medical information was the main objective of residents' routine communication with parents.
"We (residents) don't aim to give knowledge to them (parents Collaboration skills The participating residents identi ed collaboration as intra-professional health care team or working among physician colleagues. Because most patients were chronic complicated cases so they were treated with multiple specialties. Residents responded for primary health care of these patients. Treatment of speci c diseases or conditions were provided by specialists. Residents discussed and consulted with senior physicians or staffs for patient management. Residents thought parents could assess collaboration skills of residents from: (1) providing identical information; and (2) informing consultation process.

Residents' providing identical information
Parents might not see residents' consultation in front of them. The residents frequently consulted senior residents, fellows or staffs by phone calls or meetings. The parents knew whether residents effectively worked with others as a team if everyone in the team gave the identical medical information such as plans of treatment and follow up.
"….. Most of them (parents) know that we have already discussed and made plans for the patients. When specialty teams come to see the patients and inform the same plans as we told them." [FG1] Residents' informing consultation process Residents informed parents that they did made consultation or had meetings with the team. The parents often were told that residents were going to work with several health professions for making their children better. Some parents favored residents who consulted specialists they would like to have.
"Sometimes, parents are overanxious. They like us to consult too many specialists although the patients' condition is benign. If some residents do not respond to their needs, they may feel uncomfortable with the residents." [FG2] Leadership skills Residents indicated that there were several levels of physicians in a team. Commonly, one team consisted of 7-8 medical students, 1-2 junior residents, a senior resident and a staff in a general pediatric ward. Therefore, it might be di cult for parents to assess leadership skills of residents taking care of their children: "It is di cult to assess ….. There are many doctors as well as students. Maybe they (parents) can't identify whose jobs are ….." The participating residents were appreciate if the assessable skills were de ned according to their levels. Parents could assess how residents demonstrated the skills of: (1) supervising medical students or junior residents; and (2) making decision of treatment.

Residents' supervising medical students or junior residents
Junior residents (the rst year residents) lead service ward rounds and supervise and teach medical students and interns in the team. The residents demonstrated physical examination and identify important ndings to students. The parents could recognize how residents teach and advice their students.
Residents' ability to make decision or solve problems or concerns Parents of children with chronic diseases or multiple admissions would know the leader of the team.
Senior residents (the second or third year residents) helped the team in case of complicated or complex cases, di cult situations or general consultation as needed. The parents always came to senior residents if they had some major concerns. The parents expected senior residents to make decision especially on treatment plans.
"In case of unsolved problems or concerns, we (senior residents) will jump in and help our junior residents …… sometimes, they can't deal with parents or parents come forward to talk to us. I think parents can see how we can help them (parents)." [FG5] Professionalism Participants discussed how residents demonstrated their professionalism to parents in daily practice.
Residents thought parents could assess their manner, dressing or personality. Experience with in-patient admissions Children with chronic diseases mostly have had multiple admissions. The parents of these children would know the system of care team and hospital facilities. They were able to recognize students, residents, fellow or staffs. They knew the roles of junior and senior residents. They also were familiar to the treatment and medications. They would know if some residents practiced below the standard.
"They (parents) meet a number of residents so they know how we work. Multiple parents Residents indicated that parent assessment would be credible if they receive feedback from multiple parents.
"We need to collect feedback from a large number of the parents. I will doubt if I get a bad feedback from one parent but good feedback from ten parents. The one giving me the bad feedback may be unsatis ed to other aspects, not me." [FG2] Residents' perceptions of parent personal behavior Appropriate parents' expectation Each parent had different levels of expectation to the service and treatment outcome. Our center is a government hospital so the environment and service might not be luxurious as private hospital. Some parents who experienced using private hospital service might have high expectation. If they felt disappointed to the service, they could give residents' imprecise feedback.
"It is not about education. It is about attitude or biases. We can get unreliable feedback if they (parents) have bad attitude to us, hospital or anything." [FG3] Parents tended to be satis ed if they got what they wanted. They would preferred giving good feedback to residents who followed their needs.
"….. They (parents) will like residents who can give them everything or medications they want. In contrast, they unlike residents who try to explain for 15 minutes why their children don't need the medications." [FG3] Stable parents' emotional background Parents' personality or emotional background could affect residents' feedback. There was a high chance to have con ict with parents having high anxiety or over concern.
"…… If we get a negative feedback which we don't agree with, we may need to look at assessors. They (parents) may have high anxiety making them feel uncomfortable. For example, they are not satis ed with the treatment offered to their children." [FG2] Parents' confrontation to medical situations Parents in di cult situations or worsening outcome of treatment Frequently, parents confronted with di cult situations such as having children in crisis health conditions or unexpected treatment outcome. These parents might not be mentally ready to assess residents.
" ….. In acute crisis such as having fever with convulsion, the patient has seizure and come in emergency department. The parents would be very anxious. They would be frustrated or unsatis ed if we (residents) reassure and discharge the patient back to home. It is not a good time asking them to assess us." [FG3]

Discussion
Parent perception on residents' assessable skills In this study, parent perception was explored using the PARENTS questionnaire and subsequently the think-aloud interview. The PARENTS questionnaires was selected to use for this study because it was speci cally developed for pediatric residents and aimed to assess the integration skills of communication, collaboration, and professionalism [8]. We used the questionnaire to measure the con dence of parents in answering the questions. From the questionnaire results, more than 80% of the parents answered 4 or 5. This data implied that Thai parents felt con dent to assess the non-technical skills of the pediatric residents. Although there were several groups of the parents in this study (Table 2), we did not nd any differences in overall con dence levels. The educational levels or admission experiences could not predict the con dence of assessing residents' skills. Using clear questions may help the parents to clearly understand and answer each item. In a previous study, the parent demographic data did not relate to their attitude to rating student's skills in a pediatric ambulatory setting [19].
We observed some discrepancies of the percentages of the parents' con dent levels between 17 questions. More than 40% of the parents answered level 5 for questions 3 and 9. These 2 questions were related to close observation of residents' behavior. The parents clearly stated that they easily could observe residents' hand washing because the alcohol gel bottles were provided at bedsides (question 3).
The parents stayed with their child all the time therefore they could observed the interaction between the residents and the patients (question 5). More than 10% of the parents gave level 3 for questions 4, 10, 13, and 15. These questions asked how well the residents communicated with the parents and their child. However, the content for communication of these questions connected to medical knowledge or knowledge of patients' background. Some parents might not feel highly con dent to assess these aspects within the questions.
The think-aloud method allowed the parents explained the processes they used to reach the answer of each item. Our parents expressed the direct experience with the residents' skills described in the questionnaire items. They could explained why they gave either positive or negative response. The information received from our parents' responses support the assumption that parents are appropriate assessors because they are always besides their children and able to observe residents' performance [8,9]. The parents also demonstrated that their families had long-term relationships with the health care team. They recognized the team members and each member's roles. Increasing patient's understanding of the roles and training levels of physicians caring for them may impact on patient's satisfaction because they will have accurate expectation to each individual [22,23]. Our parents could identi ed the residents as their primary physicians. The parents were mostly satis ed with the residents who took care of their child.
Although there is a wide gap between doctors and patients in our culture, the parents in this study revealed that they were able to assess these residents' skills. Several reasons may explain our ndings.
First, the participating parents were familiar to the residents because more than half of them had inpatient experience more than 10 times. They encountered a number of residents and other medical staffs.
They could recognize residents' activities and compare the abilities between residents. Second, the nontechnical skills identi ed in the questionnaires were obviously seen by the parents. The parents did not have to have any speci c knowledge to assess these skills. The residents used these interpersonal skills for taking care of their patients daily. Third, it is possible that the characteristics of Hofstede's culture dimensions of Thais may gradually change due to the effects of globalization [12]. Currently, the power distance may be not very large so people feel more equal. The evidence of this nding was demonstrated in Egyptian culture. The power distance index decreased from 80 in mid-1990 to 29.6 in 2011 in Egyptian context [22]. However, this assumption need to be further investigated in other cultures.

Resident perception on assessable skills
The participating residents perceived that the parents were able to assess several physicians' skills including communication, collaboration, leadership and professionalism.
Patient-doctor communication is a vital component of health care so communication skill is the main physicians' skills assessed by patients and parents [4,5,9]. Physicians and medical trainees accepted that patients are capable of giving feedback on their communication skills. Patient feedback is positively accepted and used for physicians' improvement of performance [9]. In this study, communication skill was the rst parents' assessable skill which the residents presented to the groups. The residents proposed that parents could assess their communication skills in terms of abilities to make parents understanding the medical information, residents' active listening skills and residents' empathy expression. These are components for being a good communicator as described by CanMEDS framework [23]. However, some components of communication skill may vary according to the different workplaces.
In the pediatric emergency setting, the assessable communication skills included residents' abilities to introduce themselves, use of appropriate language, active listening skills, keep parents informed about their child's care and translate medical information to children and their parents [7].
The second assessable skill perceived by the residents was collaboration. There were two mentioned components of this skill: residents' providing identical information and informing consultation process.
The data implied that the participating residents might think of the results of the collaboration process.
The collaboration requires trusted relationship, respect and sharing of decision-making among the care team members [23]. Our residents only paid attention to the de nite plans of treatment and how the parents could receive these information. In our setting, the patients had their own sub-specialized physician as their primary physician so the patients and parents expected that the residents would consult their primary physician every visits or admissions. Therefore, the residents might try to identify the concrete actions showing their inter-professional collaboration. However, the abilities to collaborate with parents were not proposed in our residents' focus groups. This abilities was considered to be assessable by parents in the emergency care setting [10]. Having the parents and patients as a part of the teams would improve doctor-patient cooperation and make better family satisfaction [24]. The large power distance culture may in uence the residents' beliefs that physicians should work and make decision together and then give the information or conclusion to the patients.
The participating residents perceived that leadership skill was able to be assessed by the parents. The parents could judge how well the residents showed their abilities in supervising medical students or junior residents and making decision or solving the patients' problems. In in-patient setting, junior residents lead medical students for making rounds and teach medical students at bedside. Senior residents have both clinical service and academic responsibilities covering two teams in a ward. Therefore, the residents transferred the components of leadership skills from the daily working activities. As a social expectation, being a leader means abilities to demonstrate collaborative leadership and management within the health care system [23]. The leadership performance postulated by the residents might represent their collaborative leadership and management skills. However, the hierarchy in our system is also shown in the residents' perception. They considered the power of seniority in teaching junior trainees and making decision on signi cant patients' problems.
The American Board of Pediatrics provided the principles for teaching and evaluation of professionalism for residency training in pediatrics including honesty and integrity, reliability and responsibility, respect for others, compassion or empathy, self-improvement, self-awareness, communication and collaboration and altruism and advocacy [25]. However, residents' understanding of professionalism was extensively various. The common de nition of professionalism given by residents was respect, compassion, empathy or integrity [26]. In our residents' focus groups, the professionalism components were shortly discussed. The residents' perception of the assessable components of professionalism focused on the appearance of the residents. The politeness and reliability were also considered to be the assessable aspects of professionalism in this study. Our resident seemed to think of the obvious characteristics seen by parents. In a comparison study of the professionalism expectations, the important attributes of professionalism prioritized by residents and patient families were not identical although some of them overlap [27]. Patients valued communication skills and compassion more important than appearance and acknowledgement of their family members [28]. However, the parent perception of residents' professionalism was not directly asked in this study. We could not compare the similarity or distinction of parent and resident perception on professionalism.

Resident perception on credible parent characteristics
Although parent feedback as a part of MSF should be formative, the credibility and speci city of the feedback are crucial components. Trainees will use the feedback for guiding their learning and practice improvement if they perceive the feedback is accurate and comes from credible sources [10,29,30]. In this study, our participating residents revealed several characteristics causing either positive or negative credibility.
Firstly, the residents perceived that parents' experience in caring their child was important. The primary caregivers were the persons who the residents trusted that they had direct observation on the residents' skills. Feedback is more likely to be used when physicians perceive that the assessors are familiar to their work and observe their practice [33][34][35]. The experience with in-patient admissions was considered to be a signi cant factor for being credible in our residents' opinion. Because of the existence of multiple levels of trainees in the program, the parents would recognize each trainee's roles if they have had several admission experience. Residents accept the feedback from the assessors who know their roles and responsibilities [34]. Increasing opportunities for the parents to make relationship with the residents was another positive component for the credibility. Resident tend to trust the feedback coming from the patients they have longitudinal relationship with [11]. The relationship needed an adequate interaction time between the residents and parents. The parents might see the residents in different aspects during the different situations. In addition, the residents mentioned that they would like to trust the feedback from multiple parents representing multiple perspectives for high reliability and validity [11,35].
Secondly, the residents paid attention on the parents' personal behavior affecting the assessment. The parents' mental background and expectation were considered to be the signi cant components of parents' characteristics. The residents did not think the parents with high anxiety or over concern would give them the credible feedback. In a previous study, residents would like to disregard the feedback from assessors who seem uneasy or to have poor interpersonal skills [11]. In addition, feedback may be in uenced by patient's diagnosis, hospital system issues or team decision making [11]. This usually occurs when the outcomes do not meet the families' expectation. As a result, the parents may not correctly assess the residents based on their own skills. Parent standards would be vary according to several factors such as prior experience, expectation and emotional background [36]. This is also a reason to use a large number of parents to provide reliable feedback.
Lastly, the residents considered that parents facing with di cult medical situations might not be appropriate assessors. Their mental status would not be ready for assessing residents because they de nitely focused on their child's conditions. A study of parent feedback in a pediatric emergency department excluded parents who had a child requiring resuscitation [3]. Obviously, the parents would be stressful if their child's life is in crisis. Therefore, the appropriate time or situation to request the parent feedback should be speci cally considered in individual setting.

Limitation
There are several limitations in this study. This study was conducted in a single residency training program. Our institute contains speci c patient types and resident's training activities. These conditions may be not similar to other institutes in Thailand so it may not represent the perception of Thai parents and pediatric residents. Multicenter study may increase the generalizability of the ndings. Language barrier is one of the important issues of this study. There are several translation processes occurring in this research. The PARENTS questionnaires are translated from English to Thai. The qualitative data of parents and residents perception are translated from Thai to English. Some meanings of words or sentences may not be perfectly matched. Some miscommunication is undeniable.

Conclusions
Involving parents in assessment of residents' skills is challenging. This study is an initial research exploring the perception of parents and residents regarding the issues in parent feedback in Thai culture.
The participating parents were con dent to give resident feedback on communication, collaboration and professionalism. The residents recognized several parents' assessable skills including communication, collaboration, leadership and professionalism (Research question 1). The credible parent characteristics were being a primary caregiver, having admission experience and adequate contact time with the residents. Parents with high expectation, narrow-minded, emotional instability and being in di cult medical situation may provide biased feedback. Multiple parents were deemed needed to make reliable feedback (Research question 2). The data of this study may urge researchers to investigate how parents can be involved in pediatric resident assessment in their own cultures. Authors' contributions SP was responsible for the concept and design of the study, data collection, data analysis and interpretation, and manuscript drafting. JD contributed to the study design, data analysis and interpretation, and critical revision of the manuscript. PP contributed to the study design, data collection, data analysis and interpretation. All authors read and approved the nal manuscript.