Correlates of Picuriste Use in a Sample of Health-Seeking Haitian Immigrants and Adult Children of Immigrants in Miami–Dade County, Florida

Abstract

Objectives. We explored covariates of the use of picuristes (traditional health workers with no formal medical training who provide intramuscular, subcutaneous, and intravenous injections, typically with nonsterile needles) in the Haitian community of Miami–Dade County, Florida.

Methods. We surveyed a community-based sample of 205 Haitian immigrants and adult children of Haitian immigrants. Through logistic regression analysis, we sought to corroborate the correlates of picuriste use identified in previous qualitative interviews of picuristes and their clients.

Results. Picuriste injections had been obtained by 17.6% of our respondents. After control for demographic characteristics, we found that participants who reported that a trusted person recommended a picuriste were 3.9 times as likely as participants without a recommendation to have used a picuriste. Similarly, participants who believed that the benefits associated with picuriste use were worth any resulting problems were 4.5 times as likely as those without this belief to have patronized a picuriste.

Conclusions. A significant minority of our sample patronized picuristes. Our data identified factors associated with picuriste use and shed light on a frequently hidden cultural health behavior.

Cultural health beliefs and related behaviors among racial and ethnic minorities are often hidden from mainstream health professionals and may hinder efforts to eliminate health disparities in the United States. We explored a hidden health practice among Haitian immigrants and adult children of Haitian immigrants in the United States: injections provided by picuristes (injectionists). Ethnographic and clinical research on picuriste practice in Haiti defines picuristes as individuals who have no formal medical training but who provide intramuscular, subcutaneous, and intravenous injections on demand; they and their clients typically believe that these injections work faster than pills. These studies also report that picuristes routinely recycle discarded needles and syringes (C. Clérismé, Professor of Sociology and Social Research, State University of Haiti, Port-au-Prince, unpublished report, 2004, available from clerisme@yahoo.com).¹

In a study of patients diagnosed with AIDS in Haiti, Pape et al. identified picuriste injections as a potential risk factor in the progression from HIV to AIDS.¹ Common risk factors for HIV, such as intravenous drug use, unsafe blood transfusions, and unprotected homosexual intercourse were not reported as risk factors by their study participants. However, AIDS patients reported receiving more frequent picuriste injections prior to the onset of their diagnosis than did healthy control participants, and the researchers found a significant difference in the use of injected substances between AIDS patients and members of a control group. More than 94% of 36 AIDS patients reported receiving injections prior to the onset of HIV/AIDS; 58% of control group respondents reported receiving such injections. Of 10 AIDS patients with no known HIV risk factors, all had received picuriste injections in the 5 years prior to being diagnosed with HIV/AIDS.¹

Health officials in the United States may not be completely oblivious to the Haitian practice of patronizing picuristes. In a review of the Public Health Service's response to AIDS, conducted by the US Congress Office of Technology Assessment, the section contributed by the National Institute on Drug Abuse noted the use of reused and unsterilized needles by picuristes:

These cultural practices have accompanied the Haitians who have migrated to the large urban centers in the United States, and may represent a mode of parenteral exposure to HTLV-III [an early name for HIV].^2(p87)

However, no research or policy has addressed picuriste injections in the United States and the risks they pose. No correlates of lay injectionist use among Haitian or other immigrants residing in the United States have been identified, although evidence shows that other immigrants in the United States also patronize informal injectionists.^3,4

Understanding correlates of picuriste injections is important because the United States has many Haitian immigrants.⁵ The number of Haitians living in the United States is difficult to determine because many fear deportation and therefore do not participate in the census. Others may identify themselves as West Indian, Caribbean or non-Hispanic Black. The foreign-born Haitian population in Florida grew 49% between the 1990 and 2000 censuses, from 81 837 to 166 778 persons. Florida now surpasses New York as the primary place of settlement for Haitian immigrants,⁵ with the largest concentration in Miami–Dade County, where more than 109 000 persons reside, or roughly 4.7% of Florida's total population.^6,7

Understanding correlates of picuriste injections is also important because some US immigrants from other countries are accustomed to buying medications without a doctor's prescription and are familiar with injectionist practices in their native countries. They may continue using these services in the United States. An early study of the accessibility and use of unregulated pharmaceutical products, conducted in El Salvador, found that informal health practitioners were integrating medical substances into their traditional health practices, leading to possible dependence on such substances.⁸ Eighteen years later, researchers estimated that 23% to 45% of injections administered yearly in developing countries were given by injectionists without formal medical training.⁹ Injectionists practice in Thailand (where they are known as needle curers),⁹ Uganda, Pakistan, and other countries.^10,11

An anthropological review of injectionist practice observed, “Injections have been introduced into cultural contexts where other explanatory models of illness prevail.”^12(p136) Injectionists integrate injections with the plant remedies and religious rituals that are part of indigenous disease explanatory systems.^13–17 In Uganda, researchers have studied associations between the high HIV rate and unsafe injections procured in medical settings but have not found a direct causal link with the use of nonprofessional injectionists. However, lay injectionists' practices in other developing countries have been linked to hepatitis B, hepatitis C, tuberculosis, and malaria.^9,13,18,19

Picuristes have been observed to routinely recycle discarded needles and syringes (C. Clérismé, unpublished report, 2004), so risks to their patrons may be similar to the risks associated with intravenous drug use. In the United States and globally, fewer than 1% of HIV infections among Haitians are attributed to intravenous drug use.^20–22 Questions about injections from picuristes are still not asked in health surveillance surveys, perhaps because picuriste injections are not associated with intravenous drug use and because Haitian immigrants in the United States keep this practice hidden from mainstream society and conventional health practitioners. Consequently, the relative contribution of picuriste injections to health disparities for Haitians remains obscure.

Understanding correlates of picuriste injections may be a first step in identifying why the Florida Haitian community has suffered a disproportionate rate of preventable infectious disease that is not explained by commonly accepted risk factors. More than 60% of HIV-positive Haitians were classified by the Florida Department of Health as having no identified risk; fewer than 10% of HIV-positive Whites, Blacks, and Hispanics had no identified risk.^23,24 The Florida Department of Health concluded,

Compared with the total number of foreign born cases, Haitians have the largest number of HIV/AIDS cases … in any year at least 4 times greater than black foreign born cases from other countries. … With the Haitian population standing at 1% of the total population in Florida and approximately 0.05% of the US population, but making up 19% of the HIV/AIDS cases, much effort still needs to be done to decrease prevalence in these communities.^21(p9)

We conducted a community survey as part of an exploratory mixed-methods study examining the practice and use of picuriste injections among Haitians living in Miami–Dade County. In this study, we aimed to corroborate associations between reported picuriste use and demographic characteristics and between reported picuriste use and health beliefs concerning the relative value and risk of picuriste injections previously identified in interviews with small samples of picturistes and their clients.²⁵ We also sought to discover whether factors that appeared to predict the decision to obtain a picuriste injection would emerge as covariants of use in a logistic regression analysis. We hope our findings will draw attention to this relatively hidden Haitian health practice and shed light on its potential to contribute to Haitian health disparities.

METHODS

We developed a community survey from the results of a previous qualitative study in which we interviewed samples of picuristes (n = 10) and picuriste clients (n = 25).²⁵ For the present study, conducted between June 2006 and January 2007, we surveyed a community-based sample of 205 Haitian residents of Miami–Dade County.

Survey Questions

Our survey consisted of 20 questions, including demographic data such as age, gender, place of birth, length of time in the United States if foreign born, and education. Other questions were derived from responses obtained in the qualitative interviews conducted previously and elicited yes or no responses, except when the respondent refused to give either answer but insisted, “it depends.”

To obtain a descriptive estimate of picuriste use in the United States, we asked respondents whether they had used a picuriste in this country. Because word-of-mouth referral by persons who shared close blood and social ties was universal among the picuristes and clients of picuristes whom we interviewed previously, we asked survey respondents whether they had ever been recommended or referred to a picuriste in the United States. To determine whether respondents had been exposed to mainstream medical messages concerning injection safety, we asked whether they had ever heard that receiving injections from a nonmedical person is not a good idea. To assess the relative persistence of beliefs associated with picuriste practice and the potential effectiveness of future health messages regarding its use, we asked whether they would use a picuriste even if told not to. To obtain a descriptive estimate of consensus concerning picuriste use, we asked respondents to indicate whether they believed that picuristes have a place and purpose within their community. Finally, because our previous interviewees indicated that they had experienced or known of outcomes of fever, abscesses, and, on rare occasions, death as a result of picuriste injections, we also asked survey respondents whether they believed that the benefits of injections are worth any problems they could cause. (Greater detail on all questions and their derivation is available from G. J. R.)

Data Collection

We recruited 205 participants through purposive sampling procedures, including venue-based sampling, snowball sampling, and convenience sampling. Purposive sampling techniques are particularly useful for recruiting members of hidden populations.^26,27 Recruitment was initiated through leaders of Haitian community-based organizations, who provided referrals to respondents, who in turn gave names of other potential respondents. We attempted to recruit a sample with variety in age, education level, and religion. Participants were eligible if they were Miami–Dade County residents aged 18 years or older and had at least 1 parent born in Haiti.

In our previous interviews of picuristes and clients of picuristes, we found that the concept of rekonnèt (literally, to recognize) is important in the Haitian community.²⁵ It can refer to recognition that a person possesses a particular skill (e.g., picuriste injection) and typically connotes a relationship of reciprocal trust between individuals. One survey respondent assisted us by linking us with a barbershop owner, a man that she said she rekonnèt. We received permission to recruit the barber's male clients, which increased the gender balance in our sample. We also recruited at churches and college campuses to increase the variability in our sample. We did not recruit respondents by approaching them on the street.

Potential survey respondents were asked if they were willing to participate in a study examining the health practices of Haitians in Miami–Dade County, and if so, in what language they wished to respond (surveys were available in Haitian Kreyòl, French, and English). To avoid embarrassing participants who might be preliterate, we offered 2 options: the researcher could read the questions aloud and write down the responses, or the participant could read the survey and write the answers. When the participant chose the first option, the researcher read the responses to each question back to the respondent to confirm accuracy.

Statistical Analyses

We conducted bivariate analysis to explore associations between picuriste use and participant characteristics—gender, age, education, length of time in the United States, recommendation by a relative or friend, picuriste use despite being told not to, having heard that receiving injections from nonmedical personnel is a bad idea, and belief that the benefits of injections are worth any resulting problems. We then explored to what degree patronage of a picuriste was a function of the linear combination of the several characteristics identified. Because of the binary nature of reported picuriste use, we employed logistic regressions.

Our multiple logistic regression model incorporated the demographic variables gender, age, education, and length of time in the United States and respondent characteristics identified as significant in the bivariate analyses: recommendation from a friend or relative, likelihood of using a picuriste despite being told not to, and belief that the benefits of injections are worth any resulting problems. We treated age (mean = 34.9; SD = 14.7) and education (mean = 4.6; SD = 2.0) as continuous variables.

RESULTS

Characteristics of survey respondents are shown in Table 1. All respondents had 2 Haitian-born parents, 56.1% were women, and 24.9% were US born. More than 17% of respondents indicated they had obtained a picuriste injection in the United States, and 46% believed that there is a place and purpose in Miami–Dade County for those who practice as picuristes. More than half (55.6%) of respondents had not heard that receiving an injection from a person without medical training is not a good idea.

TABLE 1.

Sample Characteristics of Haitian Immigrants (N = 205) in Miami–Dade County, FL: 2006–2007

Survey Questions and Answers	%
Gender
Men	43.9
Women	56.1
Age, y
18–25	31.7
26–35	30.2
36–49	21.0
50–99	17.1
Education
0–8 y	28.3
9 y–high school graduate	28.3
> high school	43.4
Length of time in United States
≤ 5 y	17.1
> 5 y	58.0
US born	24.9
Picuriste client ever in United States
Yes	17.1
No	82.9
Would patronize picuriste if needed
Yes	36.6
No	60.0
Depends	3.4
Friend or relative recommended picuriste use
Yes	20.0
No	80.0
Would use picuriste even if told not to
Yes	22.4
No	70.2
Have heard that using nonmedical person for injection is not a good idea
Yes	43.4
No	55.6
Believe benefits of picuriste use are worth any problems that they could cause
Yes	39.0
No	56.7
Depends/don't know/refused to answer	4.3

Note. For questions to which very few respondents answered “depends,” this category is not shown.

Our bivariate analyses identified several characteristics that differentiated picuriste clients from nonclients (Table 2). We found a significant difference between clients and nonclients for education (χ² = 11.7; df = 2; P < .001). Among respondents who reported having more than a high school education, 7.9% (n = 7) had patronized a picuriste and 92.1% had not (n = 82). The length of time respondents were in the United States was inversely associated with an affirmative response to picuriste use in the United States (χ² = 6.9; df = 2; P = .03). Among immigrants, 25.7% (n = 9) of respondents who had been in the United States 5 years or less and 20.2% (n = 24) of those who had been in the country a longer time reported picuriste use; only 5.9% of US-born respondents (n = 3) were picuriste clients.

TABLE 2.

Bivariate Analysis of Survey Data From Haitian Immigrants (N = 205) in Miami–Dade County, FL: 2006–2007

	Picuriste Clients, %	Nonclients, %	χ2	df	P
Gender			0.66	1	.417
Male	20.0	80.0
Female	15.7	84.3
Age, y			2.96	3	.398
18–25	13.8	86.2
26–35	14.5	85.2
36–49	20.9	79.1
50–99	25.7	74.3
Education			11.70	2	.003
0–8 y	20.7	79.3
9 y–high school graduate	29.3	70.7
> high school	7.9	92.1
Length of time in United States			6.97	2	.031
≤ 5 y	25.7	74.3
> 5 y	20.2	79.8
US born	5.9	94.1
Friend or relative recommended a picuriste			38.39	1	<.001
Yes	50.0	50.0
No	9.2	90.6
Would use picuriste even if told not to			38.26	1	<.001
Yes	45.7	54.3
No	9.4	90.6
Have heard that using nonmedical person for injection is not a good idea			0.38	1	.535
Yes	19.1	80.9
No	16.4	83.6
Believe benefits of picuriste use are worth any problems that they could cause			30.35	1	<.001
Yes	36.3	63.7
No	5.6	94.4

Note. For questions to which very few respondents answered “depends,” so this category is not shown.

Among picuriste clients, 50% (n = 21) indicated that a relative, close friend, or significant other recommended using a picuriste. Only 9.2% (n = 15) of respondents who had not received a recommendation patronized picuristes (χ² = 38.39; df = 1; P < .001). Picuriste clients were more likely than nonclients to report that they would patronize a picuriste even if they were told not to (χ² = 38.3; df = 1; P < .001). Clients were more likely than nonclients to believe that the benefits of injections are worth any problems that could result (χ² = 30.35; df = 1; P < .001). Among nonclients, however, 63.7% (n = 51) indicated that they thought the benefits of picuriste use outweighed any resulting problems. We did find an association between having heard that obtaining an injection from a nonmedical person is not a good idea and picuriste use.

The summary measure of model fit indicated that our multiple logistic regression model was acceptable (χ² = 46.5; df = 7; P < .001; Table 3). We did not find significant differences between picuriste clients and nonclients for any of the demographic variables. Participants who reported that they would patronize a picuriste even if told not to were more than 5 times as likely as were those who answered no to this question to be picuriste users (odds ratio [OR] = 5.34; 95% confidence interval [CI] = 1.45, 19.67). Participants who had received a recommendation from a relative or friend were almost 4 times as likely as those who had not received such a recommendation to be picuriste clients (OR = 3.95; 95% CI = 1.19, 13.10). Participants who indicated that the benefits of picuriste injections are worth any resulting problems were more than 4 times as likely as those who did not share this belief to be picuriste clients (OR = 4.47; 95% CI = 1.18, 16.87). Taken together, our findings indicate that these factors were moderate predictors of picuriste use (R²_Nagelkerke = 0.46).

TABLE 3.

Multiple Logistic Regression Analysis of Survey Data From Haitian Immigrants in Miami–Dade County, FL: 2006–2007

Predictors (Reference Category)	b	SE	OR (95% CI)	P
Gender (women)	0.58	0.61	1.78 (0.54, 5.88)	.34
Age (years)	−0.02	0.02	0.98 (0.95, 1.02)	.41
Education (years)	−0.09	0.13	0.91 (0.71, 1.18)	.49
Time in US (< = 5 y)	−0.17	0.65	0.84 (0.24, 3.02)	.79
Would you use if told not to (yes)	1.67	0.67	5.34 (1.45, 19.67)	.01
Relative or friend recommended a picuriste (yes)	1.37	0.61	3.95 (1.19, 13.10)	.02
Benefits of picuriste use worth any problems caused (yes)	1.50	0.68	4.47 (1.18, 16.87)	.03
Model fit
Constant	−2.63	1.16	0.07	.02
Omnibus test that model coefficients = 0	χ2 = 46.5; df = 7; P < .001
Generalized R2	Cox & Snell = 0.297; Nagelkerke = 0.461

Note. b = unstandardized parameter estimates; OR = odds ratio; CI = confidence interval. N = 205.

DISCUSSION

We explored some of the correlates of picuriste use in a nonprobability sample of members of the Haitian community of Miami–Dade County. We found that culturally specific health beliefs regarding the trust of relatives and friends who recommend health practices and the perceived benefits of picuriste injections, relative to knowledge of health risks related to obtaining injections from a nonmedical person, were important correlates of picuriste use and that demographic characteristics were not. Patronizing picuristes even when warned against it and believing that the benefits of picuriste injections outweigh possible risks were strong predictors of picuriste use. Our findings reflect strongly held beliefs that may affect the decision to patronize a picuriste. They should be considered in the context of traditional Haitian beliefs that the quality of one's blood is a diagnostic indicator of health or illness,^28–30 that pain affects the entire body and should be eradicated swiftly, and that healers should be approachable and trustworthy.²⁰

A recommendation from a friend or relative was a strong correlate of picuriste use, a finding that is consistent with results from our previous ethnographic study, which found that rekonnèt relationships are important in the health decisions and practices of Haitians.²⁵ In the traditional Haitian health belief system, recommendation of a picuriste implies a relationship that is rekonnèt; the trust embodied in this type of social or familial relationship strongly influences choice of treatment in this culture.

Our survey respondents' strongly held beliefs about the role of picuristes and the efficacy of injections, even in the face of warnings about their risks, indicate that picuristes and their clients hold more traditional Haitian views of illness. Some US Haitians may not fully comprehend the risks involved in receiving injections from persons who are rekonnèt as healers but who lack formal medical training. Culturally relevant educational outreach about health seeking and disease prevention should be developed for this community. Such efforts should acknowledge the importance that injections and the role of picuristes have in Haitian health beliefs and describe established standards for safe injections and the consequences of unsafe injections. Because rekonnèt is an important aspect of health decisions for Haitians, practitioners and policymakers should engage individuals and community-based organizations who are likely to share the rekonnèt relationship with local Haitian residents in tailoring and delivering interventions. Also useful would be to incorporate questions about picuristes into health service assessment and disease surveillance protocols, if not nationally, then at least in geographic areas with high concentrations of Haitians.

Conventional health providers should also be educated about the practice and use of picuristes among Haitians in the United States. These efforts should focus on the importance of recognizing and respecting the cultural context of this practice. For many Haitian immigrants, traditional and valued explanations of illness that may conflict with standard medical knowledge are deeply held beliefs; replacing picuriste use with conventional medical care will not be simple. However, a clear objective would be the reduction of health risk among Haitian immigrants who might decide to obtain picuriste injections.

Limitations

Our sampling techniques and cross-sectional design have well-known limitations. The wide confidence intervals for the odds ratios of the correlates of picuriste use suggest that the estimates were not precise and should be interpreted cautiously. The very nature of immigration concerns movement from one cultural context to another. Replacing culture-of-origin beliefs concerning the nature of illness and concomitant health practices with new health beliefs and practices is a developmental process. Cross-sectional data can only be the first step in identifying relationships between characteristics of interest.³¹

Although our study did not establish a link between infection and picuriste injections given in the absence of sterile injection protocols, it raises awareness of the practice of and the reasons for picuriste injections as a health choice among some members of the Haitian immigrant community.

Conclusions

Our results provide important preliminary data concerning the factors that are associated with the decision to patronize a picuriste. Our analysis suggests that the endorsement of a personally and socially trusted person (rekonnèt), coupled with clients' strongly held beliefs about the benefits of picuriste injections are the major influences on these health decisions.

Picuriste use is a health concern because these practitioners typically have no formal medical education or training and often do not follow sterile injection procedures. Transmission of infection via unsafe injections is a function of the prevalence of the pathogen (e.g., hepatitis B, hepatitis C, HIV) and the level of injection safety.¹³ Establishing the precise risk of contracting a blood-borne pathogen from injections by lay practitioners is the subject of ongoing research. Current estimates show that the risk is not trivial and that, for intravenous injections with reused equipment, it could be as high as 15%.³² Because we found that patronage of picuristes is far more prevalent among Haitian immigrants than among their American-born adult children, it is especially important to target these immigrants with health education efforts.

Our results provide a foundation for future epidemiological studies investigating the use of nonprofessional injectionists in relation to preventable infectious disease and health disparities. It would be useful to conduct population-based case–control studies among Haitian immigrants who are newly infected with preventable diseases that are known to be transmitted through blood and assess whether picuriste injections played a role.