The dark side of doing good: a qualitative study to explore perceptions of local healthcare providers regarding short-term surgical missions in Port-au-Prince, Haiti

## Background

In response to the increasing global burden of surgical diseases, medical providers from high-income countries are increasingly volunteering in underserved regions. Short-term surgical missions are growing in number but lack appropriate oversight. Many studies on volunteers’ perspectives have been published, but few have examined the perspectives of local healthcare providers.

## Aim

To explore perceptions of Haitian healthcare providers regarding short-term volunteer missions in order to identify positive and negative perceptions.

## Methods

A qualitative approach utilizing interpretivist epistemology and a social constructivist perspective captured perceptions of participants in their social setting. Interpretivism recognizes that perceptions are subjective and socially constructed. After purposive sampling, 17 participants were recruited. In-depth recorded interviews with assistance from local interpreters explored participants’ perceptions. Using thematic content analysis, unexpected themes emerged.

## Results

Four themes emerged: general perceptions, perceived effect on healthcare system, perceived effect on healthcare workers, and future recommendations. In general, local healthcare workers appreciate the skills and knowledge that teams impart and the improved access to surgical services for the poor. However, volunteers working independent of local teams create negative perceptions, stress local healthcare providers, and strain hospital resources. Future teams should utilize local perceptions to measure impact and effectiveness of their efforts.

## Conclusions

Perceived impact of short-term surgical missions is dependent on participant perspective. This study addresses negative perceptions and offers recommendations for before, during, and after a mission. Surgical missions that emphasize training local staff and forming long-term relationships have a greater impact on surgical capacity than do missions that focus primarily on surgical volume. Further research on local perceptions is essential to ensure that local voices continue to be heard and that altruistic efforts of volunteers do not inadvertently cause negative effects.


Back Background ground
In response to the increasing global burden of surgical diseases, medical providers from high-income countries are increasingly volunteering in underserved regions. Short-term surgical missions are growing in number but lack appropriate oversight. Many studies on volunteers' perspectives have been published, but few have examined the perspectives of local healthcare providers.

Aim Aim
To explore perceptions of Haitian healthcare providers regarding short-term volunteer missions in order to identify positive and negative perceptions.

Methods Methods
A qualitative approach utilizing interpretivist epistemology and a social constructivist perspective captured perceptions of participants in their social setting. Interpretivism recognizes that perceptions are subjective and socially constructed. After purposive sampling, 17 participants were recruited. In-depth recorded interviews with assistance from local interpreters explored participants' perceptions. Using thematic content analysis, unexpected themes emerged.

R Results esults
Four themes emerged: general perceptions, perceived effect on healthcare system, perceived effect on healthcare workers, and future recommendations. In general, local healthcare workers appreciate the skills and knowledge that teams impart and the improved access to surgical services for the poor. However, volunteers working independent of local teams create negative perceptions, stress local healthcare providers, and strain hospital resources. Future teams should utilize local perceptions to measure impact and effectiveness of their efforts.

C Conclusions onclusions
Perceived impact of short-term surgical missions is dependent on participant perspective. This study addresses negative perceptions and offers recommendations for before, during, and after a mission. Surgical missions that emphasize training local staff and forming long-term relationships have a greater impact on surgical capacity than do missions that focus primarily on surgical volume. Further research on local perceptions is essential to ensure that local voices continue to be heard and that altruistic efforts of volunteers do not inadvertently cause negative effects.

INTERNATIONAL CONTEXT
The global burden of surgical disease is a significant component of public health in low-and middle-income countries (LMICs). Resource-poor nations often lack access and surgical capacity to treat injuries. A study by the World Health Organization (WHO) found that 90% of deaths in LMICs are related to injuries. 1 Every year, 20 to 50 million people in the world are permanently disabled as a result of musculoskeletal injuries, translating to 11% of total global burden of disease and injury that is treatable or preventable by surgery. 2 The wealthiest 30% of the world population receive approximately 75% of all surgical procedures; yet, the poorest 30% T receive only 3%-5%. 3 Global health experts refer to surgery as the "neglected stepchild of global health." Short-term surgical missions (STSMs) to the developing world constitute a widespread effort that attempts to fill this void in surgical capacity.
Recognizing the unmet need for surgical services, the WHO established the Global Initiative for Emergency and Essential Surgical Care in 2005. 4 Along with growing appreciation of the role of surgery in public health, interest, and desire to participate in surgical missions has increased among healthcare workers (HCW) in developed countries.
Approximately 6000 short-term medical missions (STMMs) occur globally every year. 5 STMMs may include surgical, dental, pediatric or adult general medical care. 6 Half of all US medical schools offer formal programs in "global health," 7 and 25% of recent US medical school graduates participate in STMMs. 8 The United States and Canada send the most STMMs, perhaps related to the high cost of sending missions abroad. Honduras is the top destination for American missions, and Ethiopia and Somalia are top destinations for Canadian missions. The United Kingdom sends most medical missions to Uganda and Tanzania, while Australia's top destination is Papua New Guinea. Africa receives the greatest number of missions. The top medical condition treated in LMICs is cleft lip/palate. 9 Medical missions differ in terms of number of staff (2-90), budget (a few hundred dollars to US $39 million annually), and duration (2 days to 1 month). 5 Numerous mission teams, from faith-based to non-denominational groups such as Operation Smile spend more than US$ 250 million and send thousands of volunteers abroad yearly. 6 The perceptions of medical volunteers from wealthy countries have dominated the literature; however, little research has explored local perceptions of healthcare providers. 10 At least in theory, by supporting local healthcare providers through education and mentoring, STSMs can strengthen local healthcare systems and improve access to surgical care. 11 At worst, STSMs can negatively impact quality of care, leave bad feelings with locals and undermine the local healthcare economy.

METHODS
This study aims to understand and capture the perspectives of HCWs in Haiti. An exploratory qualitative approach supported a flexible research strategy that generated informa-tion-rich data in the field. Using semi-structured interviews, conversations were conducted privately in a natural environment to explore participant experiences, 12 which are central to cultural studies. 13 Purposive sampling permitted a selection of participants who could offer rich information and generate credible representative data. 14 Using thematic content analysis (TCA), participants' responses were compared, contrasted, and categorized into common themes, and codes were generated from the data. 15

SETTING
The study was conducted at the private, urban, faith-based Hôpital Adventiste d'Haiti (HAH) in Port-au-Prince, Haiti. This hospital was selected because it frequently hosts volunteer teams ( Figure 1). Between 2010 and 2014, HAH hosted 180 teams. 16 The 72-bed facility serves approximately 25,000 patients yearly. Despite Haiti's inadequate health infrastructure, HAH remains a nationally recognized center for complex orthopedic conditions. 17 Surgical care in Haiti is delivered through private, public, and charitable hospitals, each with its own fee structure. 18 In a 4-year period, the researcher coordinated four missions to HAH, establishing rapport with the staff. All interviews took place in the air-conditioned hospital library. Refreshments and gift bags of toiletries valued at US$5 were given to each participant after the interview. A square table allowed the interviewer and interviewee to sit opposite each other and maintain eye contact. The interpreter sat beside the participant. 19

SAMPLING APPROACH
Purposive sampling identified key stakeholders to maximize useful data collection in a limited time period. 20 Diverse professions were recruited to represent a wide range of perspectives to help answer the research question. Inclusion and exclusion criteria (T Table 1) able 1) assisted in selecting potential participants. 21 The target sample was twelve participants, but five more were interested and met inclusion criteria. After five interviews, common themes began to emerge. RECRUITMENT Recruiting participants ahead of time was challenging. The first gatekeeper posted advertisements for the study a few weeks before the researcher's arrival, but few participants reported seeing it. The second and third gatekeepers intro-duced the researcher to administrators and the director of nursing, who helped to recruit interested staff. Three interviews were conducted daily for 6 days. The final sample represents a range of occupations and ages. Interview #9 was a joint interview with two subjects; they were also inter-viewed separately. Participant demographics are detailed in T Table 2 able 2.
The dark side of doing good: a qualitative study to explore perceptions of local healthcare providers regarding short-term...
Journal of Global Health Reports T The dark side of doing good: a qualitative study to explore perceptions of local healthcare providers regarding short-term...
Journal of Global Health Reports T Interviews were face-to-face, digitally recorded and conducted in Creole, the participants' native language. Relying on interpreters complicated the social context of the interview and added another layer of values, assumptions, and experiences to the data produced. 22 On assumption that more females would participate, a female interpreter was initially recruited. Two native Haitian interpreters were used; a 28-year-old local dark-skinned male and a 70-yearold light-skinned female expat. The female translator was replaced after two interviews because the data produced was not rich, and lengthy answers were frequently reduced. Interviews conducted with the local translator produced richer participant accounts. Ethics training with both translators was conducted prior to interviews. The literature review and research objectives framed the interview guide (IG) and prevented duplication of previous studies. The IG assisted in planning, exploring, and generating rich contextual accounts of participants' experiences.
The questions were open-ended, culturally sensitive, and designed to collect similar data from all subjects but were flexible enough to vary the wording and probe responses. 23 The IG and literature review helped focus on key topics:

PILOT TESTING
The selection of two pilot studies was based on convenience and access, and they assisted in feedback of the research design. 24 They were conducted with a 30-year-old male translator and 55-year-old male cast tech. No females were available for pilots. As rich descriptive data were collected, no major changes were made to the IG. The interpreter suggested shortening a few questions and removing repetitive ones. The interaction between participant, interpreter and researcher created new insights and knowledge. Four types of questions were used: introductory, "How does it make you feel..?"; follow-up, "So you're aware the volunteers are coming?"; probing, "Can you think of an example…?"; and direct questions, "Do you think the hospital should continue volunteer missions?" 25 Key questions from the IG are listed in T Table 3 able 3, and information regarding the pilot tests is provided in T Table 4 able 4.

ETHICAL CONSIDERATIONS
The primary responsibility of the researcher is the protection of the research participants. 15 Ethics approval was granted by the University of Liverpool Ethics Committee, and permission to conduct the study was approved by the Director of International Surgery and the medical director of HAH. Obtaining authorization from the local and international directors demonstrated the researcher's desire for widespread approval and transparency. Ethical considerations were discussed with all gatekeepers, emphasizing that no one be coerced to participate. 15 Volunteers were offered the patient information sheet in English or Creole. Seventeen people agreed to participate; 15 signed informed consent forms, while two preferred verbal consent. Participants were assured of confidentiality and informed that storage of data and audio recordings would be kept in a password-protected computer for 5 years and then destroyed. All identity information was removed prior to sharing results with participants. A challenge to confidentiality and anonymity was the small hospital where research was conducted.

ANALYTIC FRAMEWORK
The same IG was used for all participants and loosely structured to promote conversation and free flow of ideas and information. 26 Ten interviews were conducted in Creole with interpreter assistance, and seven were conducted in English. All interviews were transcribed into English by the researcher and saved to a Microsoft Word file.
• General perceptions of volunteer teams • Effect of teams on HCW • Effect of teams on patient care • Effect of teams on healthcare system The dark side of doing good: a qualitative study to explore perceptions of local healthcare providers regarding short-term...

Journal of Global Health Reports
TCA was used to categorize common themes generated from the data. This method transforms data from descriptive to interpretive and provides a way for the researcher to become more intimate with the data, answer the research question, and present key findings from participants' perspective. 15 Data analysis involved systematically reading and rereading the transcripts, highlighting meaningful sections, and assigning short codes to represent text. Coded data was organized into categories based on the IG. Com-paring and contrasting each interview revealed patterns, connections, and emerging themes between participants. 27 Each interview was then summarized onto a separate page for further comparison of answers. Emergent themes, subthemes, and coding scheme after data analysis are presented in T Table 5 able 5.
The dark side of doing good: a qualitative study to explore perceptions of local healthcare providers regarding short-term...
Journal of Global Health Reports T

RESULTS
Direct quotes from participants are presented to support themes that emerged during TCA. Some results were comparable to a similar study on perceptions of HCWs in Guatemala. 8 The four main themes are: This theme captures participant (P) views of positive volunteer characteristics. Administrative and frontline workers appreciate common traits but view them through different lenses. When asked how they felt about a team's arrival, most voiced excitement, happiness, and an opportunity to learn and improve access to surgical services for the poor.
"…really happy when the volunteer groups they come because a lot of the poor people cannot afford to have surgery…they are very happy and it's helpful." (P3) Although administrators appreciate team visits as well, they view free care as a financial strain to the hospital.
"By encouraging an influx of patients…we create a very serious strain on the resources of the institution…The more patients you see, the more you spend." (P9) All participants value education and knowledge sharing. With training and skill transfer, frontline workers can improve the health of the local population.
"…there are really a high level of doctors coming, so it's an honor for me to have these kind of people and work with them, and with my practice, help the patient have a better care with their expertise. It's a good opportunity for the hospital, the Haitian population and for me as a doctor." (P12) Most participants linked increased skills with personal growth; however, administrators linked staff training with opportunities for the institution to establish high standards, prestige, and financial growth.
"I long after that day when…nurses practice a culture of quality… regardless of the resources available…I'm looking into transforming this into the best hospital for everybody…we have begun to talk about medical tourism." (P6) Haitian's prefer volunteers who speak Creole to prevent miscommunication. Many expressed a need for more interpreters for documenting in patient charts for safe follow-up of patients. One frontline worker explained a negative experience in the OR that could have been avoided if an interpreter had been present. "…because they did not speak the language and asked her for an instrument and she gave the wrong one and she was all upset." (P11) UNDESIRABLE CHARACTERISTICS Frontline workers and administrators agree that among the undesirable traits is lack of cooperation with the national team. This is perceived as disrespectful and places patients at risk. "…there is a system…not do your own thing…they are supposed to work with the national team." (P1) One participant expressed the challenge involved when a volunteer tells a local to do something one way and the local chief says to do it another way. Most participants agreed that the hospital's established system and protocols should be respected. A few frontline workers gave accounts of volunteers placing patients at risk when protocols were disregarded.
Administrative support for frontline workers was not evident. One high-level worker reported abuse and discrimination by volunteers, and the administrator's response was "be patient because they will leave soon…" An administrator offered this explanation. "We are not too good at communication…" (P17) When participants were asked to describe an unpleasant experience, most typically responded "nothing to report"; however, further probing revealed negative experiences.
One frontline worker explained how volunteers undermine the trusting relationship between patients and local staff. White volunteers are called the "blan," and when the "blan" comes, the staff cannot tell patients anything because they only believe what the "blan" says. 28 Katz explains that "blan" and "neg" divide Haitian society between "them" and "us." "Blan" not only refers to skin color; it also means "foreigner." 28 Another undesirable trait reported by frontline workers and administrators was incompetence. An administrator stated that he would prefer fewer skills and better attitude from volunteers but also acknowledged the danger presented to the institution by incompetent volunteers. A frontline worker explained how incompetence influences the relationship between them and patients.
"If you don't have skills, you are a skill terrorist; you are destroying the institution and organization." (P6) "…I like competence most and dislike incompetence… By watching, you can see…some wound needs to be done in a sterile place and they do in non-sterile way…even if I tell this is no good …the patient will not accept me…Because the Haitian patient trusts them." (P15) Volunteers who act inappropriately verbally or behaviorally exhibit poor cultural intelligence and are unable to adjust their behavior to the cultural setting or interaction. 29 Three participants reported cultural insensitivity when an American walked into a clinic room and saw Haitians on their laptops. The following quote demonstrates the need to prepare volunteers for the cultural realities of poverty to prevent bad feelings.
"He… like offends me and other people who works in the department. We had a computer and he say "I didn't know there is a computer in Haiti…In a sense, we can say this guy is a racist." (P1) A few participants spoke of limiting the number of volunteers on a mission to prevent chaos and confusion in clinic and to allow locals to get more experience. One worker highlighted the difficulties presented and reinforced the desire for closer collaboration between teams.
"…if possible is for less people to come into the operation room so they can have the space to do the job. She would prefer to work close to the doctor because she learns a lot from the doctor." (P11) • General perceptions of volunteer teams • Perceptions of the effects of the volunteer teams on the healthcare systems • Perceptions of the effects of the volunteer teams on the local HCWs • Future recommendations The dark side of doing good: a qualitative study to explore perceptions of local healthcare providers regarding short-term... Frontline worker responses focused on personal advancements and positive patient outcomes, whereas administrators emphasized improved hospital image and creating policies to improve standards of care.

Journal of Global Health Reports
"…more patient coming…It's a good thing for me as a surgeon because I will have someone to help me… help the patient have a better care with their expertise…" (P12) "…my vision is to see a true source of knowledge and appreciation of norms and standards…to serve our patients better" (P6) HOW TEAMS AFFECT HOSPITAL/HEALTHCARE SYSTEM NEGATIVELY Participants identified diverse views concerning negative impacts to the healthcare system. Although most frontline workers perceived free care positively because it increases access for the poor, administrators viewed it as a strain to hospital resources and a challenge to existing fee structures. Some frontline workers were opposed to volunteers contributing to a patient fund, whereas administrators felt volunteers and patients should contribute.
"Volunteers… think we are giving free care… they don't understand that we cannot pay the nurses, pay for the generator and buy gas. …but it's not the responsibility of each volunteer to pay for it, I am against that." (P12) "…we need a system where we ask people to contribute to their health" (P9) Several frontline workers expressed poor standards of care when visiting teams worked independently of the local team. This situation portrays a negative image to the local population, who may perceive the local staff as incompetent if volunteers ignore them. Most participants preferred knowledge transfer and skills training instead of high-volume surgery to build long-term local capacity.
"Some groups…want to do too much, and for me this is not too pleasing…if you do 30 surgeries…well it's good for the patient, but I'm going to do the follow-up…if something bad happens, I don't know what to do…if you don't teach me how to do it, then you don't help me…" (P12) A few frontline workers experienced strained relations with colleagues when volunteers ignored cultural norms and traditional hierarchies. One local worker was placed in an awkward position when a volunteer excluded his chief from a work-related decision.
"He has to respect the chief; he does not want to go behind his back, he does not want a problem." (P2) THEME

TEAMS ON LOCAL HEALTHCARE WORKERS
This theme identifies the perceived impact teams have on the local staff. Although most participants described positive interactions, a significant amount voiced negative experiences. One technician expressed gratitude for acquiring the skills to improve his livelihood during clubfoot training. "I didn't earn much money, so they trained me to become a clubfoot technician…so I thank them a lot with that." (P18) Many frontline workers shared situations that increased workday stress, such as longer workdays, increased patient load, multitasking, economic strain, miscommunication, and disruption of workplace flow. The second quote refers to the fact that when teams work late, public transportation is not available; therefore, the participant must take an expensive taxi to get home. "We stay later to document. Sometimes we would stay all day and all night" (P12) "When no teams come in, I can take tap-tap (to school), but when there's a team…I take a taxi so I can go later" (P13) Communication about a team's arrival was inconsistent among males and females. Male staff tended to know in advance when a team was arriving 77% of the time, but females only knew 33% of the time. This may suggest that advance information is not disseminated to the lower-level hospital workers, who tend to be women.
"My chief and the doctor tell… some volunteers will be coming (male participant) (P2) "It is a surprise on the day they come" (female participant) (P4) Most workers initially stated that language is not an issue; however, further probing revealed communication challenges between volunteers and local staff and interpreters and local staff. For confidentiality, Haitians prefer using professional interpreters whom they will not see again. 30 The following quote describes one worker's opinion of the relationship between workers and interpreters and illustrates the power of local interpreters.
"Even if you ask somebody to help, they refuse to tell you exactly because they think you are stupid and can't learn" (P4) Miscommunication and non-communication is stressful and can lead to patient harm. Once the volunteers depart, the locals must assume care. One frontline worker described a dangerous situation that occurs when volunteers do not document in charts.
"They don't write the name in the book…We don't have any information, and I didn't see the patient and the patient is going to have surgery and I will be in charge of the patient." (P12) Most frontline workers denied feeling marginalized, but one worker experienced discrimination from volunteers. "They are trying to put me out of the circle…I didn't know if it was a sign of discrimination probably because I'm black." (P5) A few male participants developed long-term friendships with volunteers through email. These friendships were facilitated by the fact that 90% of men and 30% of women had access to email. THEME 4: FUTURE RECOMMENDATIONS Participants offered the following recommendations for future teams. A few weeks before the mission, inform the hospital of the surgical specialties on the visiting team, allowing the local surgeon to find patients to match the volunteers' skillset. Pack all supplies needed for the week to prevent straining the hospital. Prepare lectures and training programs based on local requests. Consider less surgery and more teaching for greater impact and sustainability.
"If you come here and you do 30 surgeries… well it's The dark side of doing good: a qualitative study to explore perceptions of local healthcare providers regarding short-term...

Journal of Global Health
Reports good for the patient, but I'm going to do the follow-up… if you don't teach me how to do it then you don't help me…Some groups…want to do too much (surgery) for me this is not too pleasing." (P12) Administrators suggested that teams invite and integrate community health workers to training programs and conferences to promote good will. Do not bring expired medications on missions. Some participants believe it is a form of experimentation on Haitians.
"…some medication that expired a long time ago…If you give it to somebody and they have allergy, these are problems. It's not just an animal it's a human being." (P5) Most participants emphasized the importance of coordinating with the national team and respecting the staff and established healthcare system. Respect also implies trust and solidarity. 31 "…volunteers must understand that they are not going to impose anything." (P6) "Be cooperative. Just try to talk to the local team and ask how do we do this thing?" (P1) A few participants could not identify a team's country or name. It is the patientss right to know who is administering care to them. Identification badges with job title and team name acknowledges this right.
"She doesn't know. They just call them volunteers, they don't tell them what country they come from." (P3) One frontline worker suggested that volunteers be patient and understand that Haitians are reserved and need encouragement for engagement. "Tell them Haitian people have their own culture and sometimes that all these things can be annoying." (P1) "…this nurse doesn't know how to use the monitor; then ask her does she know how to do that…" (P12) Some participants preferred small teams for greater collaboration, with volunteers focusing on sustainability, increased dialog and knowledge exchange. Maintaining high standards of care conveys respect for patients, staff and anyone watching. "…don't do that because you are in Haiti. Then it's not good for me who is watching you, it's not good for the patient or the nurses who are watching…" (P12) One participant suggested hiring two interpreters for chart documentation, which is often missing and can contribute to poor patient follow-up and poor outcomes. Participants perceived daily briefings as beneficial for education, discussion, and social interaction. Although all participants favored continuing hosting teams, administrators cautioned that without funding, future missions would be cost prohibitive.
"Now we are in really, really, really bad financial shape. We have to find a way or it won't stay an institution." (P6)

IMPLICATIONS OF KEY FINDINGS GENERAL PERCEPTIONS OF VOLUNTEER TEAMS
Most participants expressed appreciation when volunteers worked alongside the local team, spoke Creole, and were respectful of local staff and institutional rules. These findings are consistent with Green's qualitative study in Guatemala, which described participants' frustration when volunteers work in isolation of local team. 8 Working together demonstrates mutual respect, acknowledges competence, and positively affects the community's perception of the relation-ship between the local and visiting team. Unite For Sight lists lack of collaboration as one of the worst practices in global health. 32 Some participants working in the clinic and OR experienced longer workdays, a similar finding to Mitchell's group in their East Africa study, highlighting the importance of respecting the normal work hours of local staff. 10 A few frontline workers were unaware of a team's name, country of origin, and the volunteers' occupations. DeCamp et al. found misidentification as an unexpected finding in a qualitative study of medical missions in the Dominican Republic. Knowing who is providing care demonstrates collaboration and respect. 31 Most participants recognized the value of volunteer teaching and training, a concept reinforced by Mitchell et al.'s study in Tanzania, which demonstrated that strong education and healthcare systems improve surgical capacity. 10 WHO states that education is the key to positive change. 33 Morning report, bedside teaching, formal education, and morbidity and mortality meetings are investments for long-term improvements One participant felt strongly that volunteers should possess credentials and skillsets to match need. Unite for Sight agrees that incompetent volunteers are burdensome, do more harm than good and spread false knowledge. 32 Harmful practices violate social justice and human rights. When patients trust incompetent volunteers more than competent local staff, it undermines establishing trusting relationships. Green's. 8 Guatemalan study also noted that lesseducated patients put more faith in white-skinned foreign doctors than in their own doctors.
Most participants initially responded that communication with volunteers was not an issue; however, further probing revealed examples of language barriers and miscommunication. One participant felt badly when she misunderstood a surgeons' request for an instrument; another expressed mistrust of interpreters and preferred to have direct conversations with volunteers in private.

PERCEPTIONS OF THE EFFECTS OF VOLUNTEER TEAMS ON LOCAL HEALTHCARE WORKERS
The literature does not contain studies that discuss personal and spiritual benefits to the local HCS. Many participants in the current study described personal and spiritual growth from their interaction with volunteers. Participants perceived that the clubfoot-training program provided them with skills and education to improve livelihoods and increase their blessing counts. Participants often described volunteers as "angels coming down from the sky" and asked God to bless them. DeCamp et al. acknowledge themes of God and faith in interviews conducted with care recipients. 31 However, the study could not determine if participants' beliefs were a result of the faith-based NGO or of their own spirituality.
When asked if they could think of any unpleasant experiences, most participants' answered no; yet, further probing revealed stressful situations during volunteer visits. Their initial responses may be related to (a) researcher's positionality and the participants' not wanting to express negative experiences or (b) the participant's fear of job loss. Longer workdays, language barriers, misunderstandings, and lack of documentation strain the local workers. Green also found that although most people were appreciative of teams, they acknowledged the burden that they cause. 8 The dark side of doing good: a qualitative study to explore perceptions of local healthcare providers regarding short-term...

PERCEPTIONS OF THE EFFECTS OF VOLUNTEER TEAMS ON HEALTHCARE SYSTEMS
Participants shared positive and negative perceptions of STSMs' effect on the healthcare system. Teams attract more patients, improve access to surgical services for the poor, bring prestige, and improve local capacity through training and teaching. Wilson et al. agree that training the trainer is key for long-term institutional benefit. 34 When volunteers leave complicated patients behind, they must ensure to provide the locals with the knowledge to care for the patients. Some participants noted the strain on hospital resources, complicated fee structures, and doing too much surgery as compromising relations between patients and local staff.
Participants conveyed strong opinions concerning payment for volunteer surgeries. The most common perception was that free care opens the door for the poor; however, some felt patients should pay a small fee to take responsibility for their own health. Green et al. had similar findings. 8 Free care creates a culture of dependency in which patients wait for the next group of volunteers instead of seeking local doctors or getting help from a government system that already exists.
Administrators described the significant financial strain on hospital resources. During a week, volunteers typically perform three times more surgery than do local teams and use three times the amount of electricity, water, oxygen, IV fluids, generator fuel, etc. When surgical teams take over the ORs, the hospital receives no income from private surgeons-income that pays local staff salaries.
A significant challenge for the hospital is discerning who can pay, who can pay a small amount, and who can't pay. HAH uses an NGO to locate children in need of surgery, assess their living conditions, and help them pay for surgery. Many volunteers assume that patients are poor by the way they dress and offer to help them pay for services. These scenarios can become awkward when it is difficult for the volunteers to discern who can pay. Moreover, when volunteers offer to pay for a patient in front of other patients, it undermines the healthcare system. Green et al. add that free surgery is detrimental to society, as it can take responsibility off the government and further weaken the health infrastructure. 8 Most participants believed that the hospital does not compete with local doctors because the surgeries performed are complex deformity corrections. Shrime et al. concur that STSMs should be conducted in specialized surgical centers that treat a few related surgical conditions. 35

FUTURE RECOMMENDATIONS
Responses varied to the question, "What advice would you give to future teams?" Research from previous studies supports many of the findings of this study, such as cooperate with local team, match skillset to need, be culturally sensitive, provide more training and less surgery, respect local staff and hospital rules, etc. Grimes et al. concur and add consideration of financial impact and returning yearly to build local capacity. 36 Mitchell et al. agree that volunteers' understanding of local culture prevents insensitivities. 10 Green et al suggest that recipients of short-term groups perceive benefit from free care, improved access to healthcare and specialists, exchange of experiences and knowledge, and additional supplies. 8 To prevent stress from patient overload of frontline workers Montgomery suggests that volunteers consider compensation. 37 According to most participants, volunteers who respect local staff and culture are viewed positively. Mitchell et al. agree that volunteer understanding of local culture pro-motes sensitivities to cultural norms and values. 10 Participant perspectives and their recommendations are discussed under study recommendations.

STRENGTHS OF RESEARCH
The strength of this study is the qualitative research design, which captured the experiences, views, and opinions of the local HCWs. Most accounts were rich, and participants shared their stories and personal perceptions regarding the impact of STSMs. Qualitative methods filled the gap between previous studies and experience in the field. 13 Through in-depth interviews, participants were able to reflect on both positive and negative aspects of hosting visiting medical and surgical teams.
The flexible and iterative nature of the research design presented unexpected discoveries such as hospital costs of hosting missions and negative perceptions of teams working independent of the national team. Qualitative inquiry presented contrasting views of participants regarding free versus fee for service and desirable attributes of volunteers. An advantage of the researcher's outside position is the necessity for participants to explain their answers in more detail for the researcher's understanding.
Purposive sampling permitted the selection of participants who could generate rich data and diverse experiences within the study time constraints. Because the researcher was an outsider, gatekeepers were needed to gain access to participants. Although gatekeepers have the power to add credibility and validity to the study, they can also create barriers to access. 38 The diversity of frontline workers and administrators offered divergent views for comparison. The proposal aimed for one gatekeeper to recruit six males and six females; however, staff availability, time constraints, and staff interest determined the final sample universe.
Pilot interviews confirmed the site was accessible and comfortable, and participant feedback informed fine-tuning of the IG. Pilot studies added confidence to the researcher's ability to probe participants further. The local interpreter quickly established rapport during the second pilot study. Some studies suggest that interpreters should use their own judgment when probing participants, while others argue that probes should be used only under the direction of the researcher. 39 The interpreter probed further when asked by the researcher but may also have probed independently without the researcher knowing.
Ethical approval at local and international levels conveyed wide support for the study and promoted sensitivity and compliance with ethical standards (i.e., informed consent, transparency, and confidentiality).
The language divide in Haiti was an unexpected finding. Most participants could not read the informed consent form in Creole. Haitians speak Creole, but most do not read or write it, and less than 10% understand French. French is spoken among professionals, but Creole is spoken to patients. Linguistic exclusion prevents the poor from participating in political and social processes because official documents are in French. Pierre reports that the language divide contributes to economic gaps and inequality. 40 Kapborg and Bertero suggest that the "three-way production of data" during cross-cultural and language research is complex, challenging and ultimately determines the validity and findings of the study 41 Williamson reports that gender and socio-demographic matching may enhance trustworthiness and credibility, but familiarity may limit the depth of data produced. 39 Interpreters may not relay the participants' reply if it represents their ethno-cultural group negatively. Colin's cultural training tool suggests that The dark side of doing good: a qualitative study to explore perceptions of local healthcare providers regarding short-term...

Journal of Global Health Reports
Haitians mistrust interpreters and prefer family members to maintain confidentiality. 30 However, if family is not available, they prefer professional interpreters with whom they have no relationship and will not see again. If financial constraints were not an issue, a professional interpreter would have been hired.
The interpreter explained consent forms in Creole to participants who could not read English or Creole. Flexibility in obtaining verbal or written consent conveyed acceptance of cultural norms. Participants were assured that demographic information would be deleted to protect identities before research findings would be disseminated.

LIMITATIONS OF RESEARCH
The complexities and limitations of the study are related to language barrier and cultural factors that affected the relationship between researcher, interpreter, and participant. 42 Some argue that the influence interpreters have on the production of knowledge is a limitation to cross-cultural studies, while others suggest that it adds meaning to the interpretation. 39 The small sample size from a single hospital makes findings cautiously generalizable. Few published articles exist on local perceptions; therefore, external references to compare findings are limited. 43 Because participants knew the researcher as a team coordinator, positive responses may be biased to promote the researcher's return on future missions.
Some studies suggest that gender matching improves the perception of trust and credibility among participants 39 ; therefore, a female interpreter was originally recruited. However, building rapport proved more important than gender matching. Despite ongoing training, interviews continued with long responses but short interpretations. The first interpreter was replaced with a local bilingual male employee of the hospital who provided richer accounts from participants. However, the second interpreter's powerful position and similar background with participants may have inhibited their openness. 39

LESSONS LEARNED
To increase rigor, more sites and participants would be included to compare perceptions. A shorter, more simplistic and culturally worded IG may have produced thicker descriptions in a shorter amount of time. Given that few workers could initially describe an unpleasant experience, that question would be deleted. Negative experiences were often expressed indirectly in later conversations.

RELIABILITY AND VALIDITY
A significant barrier to data dependability and credibility is the use of two interpreters who did not participate in the data analysis. An honest and clear description of data collection and analysis combined with audio recordings provided an audit trail for transparency. 44 Analysis of data is supported with direct quotes from participants, including divergent views to maximize validity of findings. Triangulation between researcher and dissertation advisor provided clarity of different perceptions. Triangulating the data or comparing findings with other investigators could produce richer accounts and increase credibility. 45 TRANSFERABILITY Providing information-rich descriptions of the setting and participant responses allows the reader to decide the relevance of the study in a broader context. 44 Burchett et al. suggest other factors that influence relevance, such as applicability of findings, congruence with beliefs and values, ease of implementation and adaptation. 46

PUBLIC HEALTH RELEVANCE
The relevance of this qualitative study underlines the importance of local perceptions in humanitarian work. Looking past Northern bias to understand the impact of actions can improve delivery of surgical services and help to evaluate the effectiveness of services provided. 47 Visiting teams that impose Western perspectives and opinions on local HCWs promote negative perceptions. Missed opportunities create barriers to engage local staff, build local capacity, and understand the broader cultural and political context of our actions. 47 Negative perceptions create relationship barriers, which affect every aspect of mission work, including brain drain from under-resourced health systems. 47 The ultimate goal of STSMs is to be put out of business by creating a sustainable healthcare system with trained, competent local staff.

FURTHER RESEARCH
This research was an exploratory study of the perceptions of HCWs regarding STSMs. Further research may include patients' perspectives to gain further insight into the care recipients' perspectives. Local healthcare views and opinions should continue to be heard through other qualitative studies to inform public health policy, enforce ethical standards, and develop an evaluation tool to assess volunteer actions.

CONCLUSIONS
As global health programs and collaborations increase, volunteers often overlook the hosts' perspectives on their efforts. Most published studies have assessed these collaborations from the volunteers' perspectives and occasionally from the patients' perspectives. Through the voices of the local healthcare personnel, this study confirms previous findings that the burden of surgical diseases can be reduced by education and training of local HCWs through long-term commitment. It also offers concrete suggestions, based on findings from qualitative interviews, to improve and re-focus volunteer efforts.
Best outcomes begin with choosing appropriate volunteers who are experts in their fields. Skills training and education are highly valued by local workers and paramount for sustainable improvements. The clubfoot program at HAH is an example of recurring volunteer efforts bringing about sustainable improvement.
Measuring impact and effectiveness of volunteer interventions requires the input of local perceptions. Developing relationships anchored in trust allows locals to express perceptions and volunteers to listen. With better understanding of local perceptions, volunteer teams can avoid creating negative perceptions and make substantive improvements.
Local HCWs have experience and insight that can help volunteers build mutually beneficial relationships, improve patient care, plan and execute short-term missions, and ultimately contribute ethically toward global health equality. Volunteers need to come to terms with the fact that volunteerism is not automatically positive and that doing good also has a dark side. Increased planning, understanding, cultural humility, and a tighter focus on teaching rather than on providing service will strengthen rather than supplant or bypass local talent. In the words of one insightful volunteer, we should "leave skills, not just scars." The dark side of doing good: a qualitative study to explore perceptions of local healthcare providers regarding short-term...

Journal of Global Health Reports
A Ackno cknowledgements: wledgements: I am grateful to the staff of Hôpital Adventiste who graciously gave their time to make this project possible. Parts of this study were submitted in fulfillment of my thesis requirement for the master's degree in international public health from University of Liverpool.
Funding: Funding: No external funding received. Authorship c Authorship contribution: ontribution: MC is the sole author (conceived of the project, collected data, analyzed the data, and wrote the manuscript for submission).
C Competing interests: ompeting interests: The author completed the Unified Competing Interest form at www.icmje.org/coi_disclosure.pdf (available upon request from the corresponding author), and declare no conflicts of interest.
C Correspondenc orrespondence to: e to: Merrill Chaus, RN, BSN, MPH Johns Hopkins School of Nursing Baltimore, Maryland USA Mchaus1@jhmi.edu The dark side of doing good: a qualitative study to explore perceptions of local healthcare providers regarding short-term...