Design and Implementation of the Puerto Rico Observational Study of Psychosocial, Environmental, and Chronic Disease Trends (PROSPECT)

Abstract

The Puerto Rico Observational Study of Psychosocial, Environmental, and Chronic Disease Trends (PROSPECT) is a prospective cohort study in Puerto Rico (PR) aiming to identify trends and longitudinal associations in risk factors for cardiovascular disease (CVD). In 2019, PROSPECT investigators started recruiting a sample of 2,000 adults aged 30–75 years in PR using multistage probabilistic sampling of households and community approaches. Culturally sensitive trained research assistants assess participants, at baseline and at 2-year follow-up, in private rooms at a network of partner clinics. The study collects comprehensive data on demographic factors, socioeconomic and environmental factors, medical history, health conditions, lifestyle behaviors, psychosocial status, and biomarkers of CVD and stress. PROSPECT will estimate the prevalence and incidence of psychosocial, lifestyle, and biological CVD risk factors, describe variations in risk factors by urbanicity (urban areas vs. rural areas) and exposure (before and after) to natural disasters, and determine predictors of longitudinal changes in CVD risk factors. The study has 4 coordinated operational strategies: 1) research productivity (including synergy with existing epidemiologic cohorts of Hispanics/Latinos for comparison); 2) research infrastructure (biorepository, ancillary studies, and clinical research network); 3) capacity-building, education, and training; and 4) community outreach, dissemination, and policy. PROSPECT will inform public health priorities to help reduce CVD in PR.

Abbreviations

COVID-19

coronavirus disease 2019

CVD

cardiovascular disease

PR

Puerto Rico

PROSPECT

Puerto Rico Observational Study of Psychosocial, Environmental, and Chronic Disease Trends

PRADLAD

Puerto Rico Assessment on Diet, Lifestyle, and Diseases

RA

research assistant

Adults living in the US island territory of Puerto Rico (PR) have high estimated prevalences of type 2 diabetes (16%), high cholesterol (37%), hypertension (45%), excess body weight (70%), and cardiovascular disease (CVD; 15%), as of 2018 (1). They also have unhealthy lifestyle behaviors, including poor diet quality, inadequate intake of several nutrients, and an 11% prevalence of tobacco use; only 6% of PR residents meet the US national guidelines for physical activity (1, 2). Few reports have documented the prevalences of psychosocial risk factors for CVD, such as depression, stress, and trauma, and available coping mechanisms in PR.

Scientific studies support a link between psychosocial, lifestyle, and environmental risk factors and eventual chronic disease (3–5), yet there is a lack of population-based research in PR on the prevalence and interconnections of CVD risk factors. The few extant epidemiologic studies either are outdated, are limited to the San Juan metropolitan area, have a cross-sectional design, have only partial data, or have nonprobabilistic or small samples. Thus, Puerto Ricans’ health remains markedly understudied at the population level, despite the clear and pressing need to alleviate their disease burden. To add to this, in September 2017 PR was devastated by 2 hurricanes (Irma and María), imposing urgency for studying risk factors for CVD under times of distress.

In response, the Puerto Rico Observational Study of Psychosocial, Environmental, and Chronic Disease Trends (PROSPECT) was established with funding from the National Heart, Lung, and Blood Institute. PROSPECT is an islandwide cohort of 2,000 adults aged 30–75 years residing in PR and recruited using a multistage probabilistic sampling approach. Its overall goal is to identify trends and longitudinal associations in psychosocial, lifestyle-related, and cardiometabolic risk factors that can guide public health priorities and future research needs aimed at reducing CVD-related disparities in PR. The project is a partnership between the Harvard T.H. Chan School of Public Health (Boston, Massachusetts), FDI Clinical Research (San Juan, PR; the core research site), and the University of Massachusetts Lowell (Lowell, Massachusetts). In this article, we describe the aims, design, data collection, and coordinated operational strategies of this novel study.

METHODS

Summary

PROSPECT is a prospective population-based cohort study consisting of a baseline examination, a 2-year follow-up examination, and annual follow-up telephone calls made to inquire about new diagnoses or significant clinical events. The study is recruiting 2,000 adults aged 30–75 years living in PR using a multistage sampling strategy. Examinations are carried out at one of more than 50 partner clinics across the island. Trained research assistants (RAs) conduct an in-person comprehensive interview to probe for sociodemographic and medical information, implement validated questionnaires, measure anthropometric characteristics, and obtain hair, urine, and fasting blood and saliva samples for laboratory analyses. Participants remain enrolled in the study for its duration, unless they are lost to follow-up. PROSPECT has approval from the institutional review boards of the Harvard T.H. Chan School of Public Health, Ponce Health Sciences University (Ponce, PR), and the University of Massachusetts Lowell. The study is registered under ClinicalTrials.gov identifier NCT03794531.

Scientific aims

PROSPECT’s analytical aims are to 1) estimate the baseline and 2-year prevalence of psychosocial, lifestyle, environmental, and biological CVD risk factors; 2) estimate incidence and trends in CVD and other major chronic diseases; 3) determine longitudinal associations between 2-year changes in psychosocial, lifestyle, or environmental factors and biological markers of CVD, as well as eventual chronic disease; and 4) estimate differences in prevalence and 2-year changes in risk factors for residents according to urbanicity of residence (urban vs. rural) and postdisaster exposures. The overall hypothesis is that adults in PR will show poor lifestyle and psychosocial health and that these qualities will be associated with unfavorable CVD risk factors. The prevalences and associations will differ between rural and urban areas and will be unfavorable after a natural disaster.

Participant sampling, recruitment, and eligibility

PROSPECT employs a multistage sampling strategy shown to capture a probabilistic and representative sample of residents in areas impacted by natural disasters (6, 7). The sampling design ensures a sufficient sample size for stratification by urbanicity. Sex-based and age-based oversampling are conducted to reach levels of 30% male participants and participants aged 45–75 years who exhibit most of the CVD risk factors. The multistage sampling procedure consists of both probability sampling and communitywide sampling. In probability sampling, we first identify potentially eligible participants using 2010 US Census block frames with socioeconomic and demographic data compiled by the Institute of Statistics of Puerto Rico. We use door-to-door household enumeration protocols implemented by previous studies inclusive of Puerto Ricans (8, 9). Blocks are visited 3–5 times, on different days of the week, and at varying times of the day or evening. Only 1 participant per qualified household is randomly invited to participate. Sampling weights are available for analysis. In communitywide sampling, we advertise the study at community locations (including partner clinics), events (e.g., health fairs, festivals), and in social and mass media via flyers and postings, and through referrals from other participants. As part of this sampling, we have invited a subset of potentially eligible adults who participated in the Puerto Rico Assessment of Diet, Lifestyles and Disease (PRADLAD), a pilot study of 380 adults conducted in 2015 (10, 11).

Up to 4 attempts are made to contact participants. Those who respond proceed to be screened for eligibility; otherwise they are deemed “unable to reach.” Interested individuals are screened using a phone-administered questionnaire for the following self-reported eligibility criteria: 1) age 30–75 years at the time of enrollment; 2) not being institutionalized at the time of enrollment; 3) living in PR at the time of enrollment and for at least the previous year, and not planning to move away from the island within the next 3 years; 4) living in a stable dwelling (owned, rented, or residing in, and without concerns about eviction) at the time of enrollment and for the prior and subsequent 2 months; and 5) being able to answer questions without assistance in Spanish or English—that is, not having any severe or untreated psychological disorder or mental/physical disability that would prevent completion of an interview (e.g., schizophrenia, aphasia, severe speech impairment, or advanced Alzheimer disease/dementia). Persons with pronounced impairments observed by RAs are excluded.

Recruitment will continue until 2,000 eligible participants have been enrolled with signed informed consent. We expect to screen 2,900 individuals at baseline to meet the goal. We estimate an 85% retention rate for visit 2; the projected 1,700 participants completing both visits will provide adequate power for testing all study aims. We record reasons for ineligibility, refusal, or loss to follow-up (including date and cause of death reported by contacts).

Examination overview

The primary-care clinics where PROSPECT study visits take place are distributed islandwide and are within 45 miles (72.4 km) or 45 minutes’ travel (by car), at most, from a given municipality (Figure 1). Partner clinics were selected on the basis of geographical coverage and to capture a diverse sociodemographic population by including primary-care clinics at research-based clinical sites, city hospitals, and community clinics encompassed by the Association for Primary Health of Puerto Rico, a nonprofit organization funded through the US Department of Health and Human Services. To date, more than 50 clinics have partnered with PROSPECT.

Figure 1.

Locations (stars) of partner clinics and hospitals where clinical procedures will be carried out and participants will be interviewed in the Puerto Rico Observational Study of Psychosocial, Environmental, and Chronic Disease Trends.

All clinic settings have state-of-the-art facilities, private rooms for interviews, adequate facilities and staff for blood draws, and easy access and familiarity for participants. Participants may request reimbursement for transportation or parking expenses; free transportation to the main study site is available (FDI Clinical Research). Appointments are scheduled at the closest participating clinic in the morning, as participants are asked to fast for 12 hours prior to the appointment. Participants are called to be reminded of their appointment and given detailed instructions on how to prepare for the visit. At this call, they are asked for their contact information, as well as information on 4 contact individuals who could locate them.

Bilingual (i.e., Spanish and English) RAs conduct all interviews and examinations except the blood draw, which is conducted by a licensed nurse. RAs undergo thorough periodic training in all standardized protocols, human subjects research considerations, safety measures, medical record abstraction, software use and data capturing, and culturally sensitive interviewing methods. Interviews and measurements take approximately 3–3.5 hours, with the following expected durations: 1) welcome and informed consent—10 minutes; 2) sociodemographic and medical history questions—20 minutes; 3) biological samples—30 minutes; 4) anthropometric measurements—15 minutes; 5) break for rest and a snack—5 minutes; 6) food frequency questionnaire—40 minutes; and 7) health behavior, psychosocial, and environmental questionnaires—1 hour and 15 minutes.

Procedures are ordered to first conduct all measurements that must be obtained in the fasting state, followed by a short break and a healthy snack before proceeding with questionnaires. Participants may pause or end the examination at any time. Depending on when the examination is interrupted, we reschedule another in-person visit or complete the interview over the phone. Participants are recontacted up to 3 times by phone within the subsequent 14 days to finish the interview. If the participant chooses not to reschedule, measurements are marked as missing data. If a participant chooses not to be recontacted, loss to follow-up is assigned. Participants receive a $50 gift card as an incentive upon signing the consent form at the baseline visit.

Detailed procedures

During the visit, assessments are made of demographic and socioeconomic characteristics, medical history and health care, biological and body measurements, health behaviors, psychosocial tests, and environmental factors (Table 1).

Table 1.

Key Data and Measures to Be Obtained at Baseline (2019–2021) and the 2-Year Follow-up (2021–2023) in the Puerto Rico Observational Study of Psychosocial, Environmental, and Chronic Disease Trends

Measure	Description
Sociodemographic factors
Demographic factors	Age, sex at birth, marital status, ethnicity, municipality of residence
Household composition	Number, ages, and relationships of people living in the household
Migration	Migration history, including duration and reasons for moving
Socioeconomic factors
Income	Household income
Occupation	Employment history
Education	Duration (years) of formal education
Medical history
Medical history	Self-reported medically diagnosed diabetes (by type), high blood pressure, dyslipidemia, angina, heart attack, heart failure, atrial fibrillation, stroke/transient ischemic attack, heart or vascular procedures, kidney disease, liver disease, hepatitis, migraines, cancer, respiratory problems, thyroid conditions, gastrointestinal conditions, depression, arthritis, osteoporosis, eye disease, physical disabilities
	Family history of major chronic diseases
Medication use	Use of prescription and over-the-counter medication, dose, duration
Health-care access	Health insurance, use of services, barriers to health care, self-rated health, and adherence to medical advice
Women’s health	Pregnancy and menopause history
Biological/body measurements
Anthropometry	Weight, height (or knee length), waist and hip circumferences, body composition through bioelectrical impedance
Blood pressure	Seated systolic and diastolic pressures and heart rate
Hair sample	Hair sample for cortisol assay
Saliva sample	Passive drool sample for storage
Fasting blood sample	Total, HDL, and LDL cholesterol, triglycerides, glucose, insulin, hemoglobin A1c, C-reactive protein, platelet count, fibrinogen, renal and liver panel, and DHEA-S
Spot urine sample	Epinephrine, norepinephrine, dopamine
Health behaviors
Tobacco use	History of tobacco use: type, amount, frequency, and secondhand exposure
Alcohol drinking	Alcohol use history: type, amount, and frequency
Sleep	Quantity and quality of sleep
Physical activity	Level and amount of physical activity and sedentary behaviors
Dietary assessment	Semiquantitative food frequency questionnaire
Food-related behaviors	Food purchases, cooking and eating habits, mealtime frequency, foods consumed away from home, drinking-water access, food security and assistance, disordered eating, diet quality, home food inventory
Psychosocial markers
Depressive symptoms	Center for Epidemiologic Studies Depression Scale
Anxiety symptoms	Generalized Anxiety Disorder Scale
Trauma symptoms	Abbreviated form of the PTSD Checklist—Civilian Version
Stress	Perceived Stress Scale and Chronic Stress Scale
Loneliness	UCLA Loneliness Scale
Discrimination	Perceived recent and lifetime discrimination questionnaire
Social support	Cohen’s ISEL-12; Norbeck Social Support Questionnaire
Familial social network	Social and community support and assistance
Coping	Brief Resilience Scale; Brief Resilient Coping Scale
Social connectedness	Social Connectedness Scale/Social Assurance Scale
Environmental factors
Neighborhood safety	Food access, walking environment, safety, violence, social cohesion, neighborhood activities
Hurricane-related exposuresa	Personal impact, property loss, loss of services and resources
Pest-related exposures	Exposure to various animal pests and contaminated water
Adverse childhood experiencesa,b	BRFSS ACE module

Abbreviations: ACE, adverse childhood experiences; BRFSS, Behavioral Risk Factor Surveillance System; DHEA-S, dehydroepiandrosterone sulfate; HDL, high density lipoprotein; ISEL-12, 12-item Interpersonal Support Evaluation List; LDL, low-density lipoprotein; PTSD, posttraumatic stress disorder; UCLA, University of California, Los Angeles.

^a Not assessed at the 2-year follow-up.

^b The BRFSS ACE module was adapted from the Adverse Childhood Experiences Study and collects information on child abuse and neglect and household challenges.

Demographic, socioeconomic, and medical history questionnaires.

RAs administer a questionnaire to ask about age, sex at birth, ethnicity, educational attainment, household income, marital status, work history, migration history, area of residence, and household composition. Medical history includes self-reported medical diagnoses of a comprehensive list of conditions, including cardiovascular procedures. For each affirmative diagnosis, we obtain detailed information on medication use, year of diagnosis, and current status of the disease. To assist with this, participants are asked before the visit to bring a list of their current medications and dosages, and RAs review the list with them. We probe for a family history of 8 major chronic diseases, self-rated health, use of health-care services, health insurance, and adherence to medical advice for major cardiometabolic conditions (10). Participants who respond “female” to the question on sex at birth are asked questions on history of pregnancies and menopause. The questions for this section were obtained from previous studies carried out among Puerto Ricans, with minor modifications for the PR context (8, 10, 12).

Biological samples.

Fasting blood samples (i.e., no food or drink other than water for 12 hours prior) are obtained via venipuncture by a licensed nurse trained following safety procedures from a standardized protocol. Plasma is separated within 2 hours onsite in a refrigerated centrifuge, and samples are refrigerated or frozen until collected by Toledo Clinical Laboratories (Arecibo, PR) on the same day as the blood draw. Toledo Clinical Laboratories is a certified laboratory facility with collection sites across PR for streamlined and centralized processing of biological samples. Samples are processed for analysis of total cholesterol, plasma triglycerides, high-density lipoprotein, low-density lipoprotein, blood glucose, insulin, hemoglobin A_1c, C-reactive protein, platelet count, fibrinogen, dehydroepiandrosterone sulfate, and a comprehensive metabolic panel with liver and renal biomarkers. Consenting participants have a blood sample stored at −80°C for potential use in future studies.

Participants are given a collection cup and detailed instructions for spot urine collection. Urine samples are collected on the same day by Toledo Clinical Laboratories and analyzed for epinephrine, norepinephrine, and dopamine. Consenting participants have a urine sample stored at −80°C for potential use in future studies. Passive drool saliva samples are obtained and stored at −80°C for future studies.

To obtain a hair sample, 1–3 cm of hair is cut with clean scissors from the posterior vertex of the skull as close to the scalp as possible (13). The hair sample is wrapped in aluminum foil and placed in a clean paper envelope. A brief questionnaire on use of steroids and hair-care procedures (i.e., coloring, bleaching, straightening, exposure to chlorine) is administered for adjustment of cortisol levels during analysis, if appropriate. Samples are shipped at room temperature to the University of Massachusetts Amherst for analysis of cortisol using standard protocols (13). Residual hair samples of consenting participants are stored at room temperature for 1 year and then at −20°C for potential use in future studies.

Anthropometric measurements.

Participants are asked to wear light clothing and to remove outerwear. While the participant is standing with feet close together, arms at their side, body weight evenly distributed, and breathing normally, RAs measure waist and hip circumferences at the end of a normal expiration using a stretch-resistant measuring tape, weight using a portable clinical-grade scale, height using a portable stadiometer (or leg length from the knee to the ankle joint, converted to predicted height, for participants unable to stand), and body composition using a bioelectrical impedance scale, following standardized protocols (14, 15). Pregnancy and use of pacemakers are exclusions for bioelectrical impedance. Blood pressure and pulse are measured at the beginning, midpoint, and end of the visit with an electronic sphygmomanometer on either arm after a 5-minute rest with the participant seated and feet flat on the floor. All measurements are repeated twice, with a third measurement taken if the first 2 differ by a predefined margin of error; the averaged value is used for analysis.

Health behaviors.

A food frequency questionnaire adapted and validated against biomarkers of intake for this population is administered to assess the frequency and quantity of intake of various foods and beverages (8, 16, 17). A model of standard portion sizes is shown to participants to help them estimate serving sizes. The Minnesota Nutrient Data System (version 5.0_35) is used for food and nutrient analyses. In addition, we ask questions on cooking and eating behaviors, mealtime frequency, foods consumed away from home, and drinking-water access and administer a brief diet-quality screener referring to intake before Hurricane María (10, 18, 19). The Three-Factor Eating Questionnaire is used to assess cognitive restraint, uncontrolled eating, and emotional eating (20). We include a culturally adapted Fulkerson Home Food Inventory to assess foods available at home before, during, and after Hurricane María (21). Household food insecurity is assessed using the US Department of Agriculture’s Household Food Security Survey Module (22, 23).

Detailed data on alcohol intake and tobacco use and exposure are collected (8). Physical activity at various levels at work, outside, or at home on weekdays and weekends is captured using a modified version of the Paffenbarger questionnaire from the Harvard Alumni Activity Survey, which has been validated against objective measures of activity and tested in Puerto Rican adults (8, 24). Questions on hours of sleep over a 24-hour period and quality of sleep are asked (25).

Psychosocial measures.

All of the psychosocial measures have been validated and implemented in Puerto Rican or other Hispanic populations. Short-term stress is measured with the Perceived Stress Scale (26–28). A chronic stress inventory is used to assess ongoing stress in several life domains related to health problems in oneself or others (29, 30). Depressive symptomatology is captured using the Center for Epidemiologic Studies Depression Scale (31–33). The abbreviated form of the PTSD Checklist—Civilian Version asks about reexperiencing, avoidance, and hyperarousal regarding past traumatic events (34, 35), and the Generalized Anxiety Disorder screener assesses anxiety symptoms (36–38). The 3-item UCLA Loneliness Scale is used to measure overall loneliness (39). Measures of perceived major discrimination and everyday discrimination capture information on acute and observable experiences of discrimination and unfair treatment across various features (40).

We include questions on extent and frequency of social ties, along with associated perceived support and assistance from these ties (27). The Interpersonal Support Evaluation List-12 assesses social support in 3 domains: appraisal, belonging, and tangible support (41, 42). The Spanish Brief Resilience Scale assesses resilience as the ability to recover from stress (43). The Spanish Brief Resilient Coping Scale measures personal perceived competence, optimism, life satisfaction, positive affect, and coping tactics (44). We assess sense of belonging within a group using the Social Connectedness/Social Assurance Scale (45).

Environmental factors.

A neighborhood safety scale measures participants’ perceived availability of food access and walking environment, esthetic quality, safety, violence, social cohesion, and activities with neighbors (46). An adapted hurricane-related exposure questionnaire asks about personal impact, property loss, and loss of services and resources after Hurricane María (47, 48). A pest-related questionnaire asks about exposure to various animal and insect pests and water pollution. The Centers for Disease Control and Prevention’s Adverse Childhood Experiences Module gauges early negative or traumatic experiences (49).

Medical record.

Clinical assessments, diagnoses, medication use, and laboratory results from the last date available are abstracted from consenting participants’ paper-based or electronic medical record and entered into their study record. We will follow protocols previously used in PRADLAD (10). If a participant is not affiliated with a partner clinic, we request access to medical records from their primary-care physician. We validate self-reported data by comparing them with medical record data.

Follow-up examinations and alternative protocols

One year after each interview, participants are recontacted by phone to update all contact information, sociodemographic data, medical diagnoses or procedures, use of medication, and any major changes in their health and behaviors. Two years after baseline, participants are asked to return for an in-person visit to repeat all biosamples and clinical measurements, update medical diagnoses and medications, and repeat the behavioral and psychological questionnaires.

Because PR is vulnerable to natural disasters, we have contingency plans for pausing or modifying protocols in case of hurricanes or major storms, earthquakes, tsunamis, droughts, or other events such as political unrest. We record details of any event that may impact the study operations or the data provided by participants. In March 2020, the coronavirus disease 2019 (COVID-19) pandemic interrupted all in-person activities as PR imposed strict prevention measures. During this time, RAs called participants to complete any unanswered baseline questions by phone or video chat or to make annual follow-up calls to those on such a schedule. Retention strategies also continued during this time. Furthermore, we implemented a weekly COVID-19 questionnaire to monitor participants for signs and symptoms, testing, management, food access, and behavioral and psychosocial factors during the pandemic. In June 2020, PR eased restrictions, and RAs now take clinical measurements and collect biological samples in person at the clinic and administer all questionnaires by phone or video chat. We provide participants with culturally relevant and trustworthy information on COVID-19 and connect them to available resources for mental and clinical services (including COVID-19 testing sites). The study team meets weekly to adjust protocols, timelines, and ethical approval, as necessary.

Timeline

PROSPECT started recruiting in January 2019, and eligible participants were invited to their baseline interviews starting in March 2019. The investigators aimed to complete recruiting of the targeted 2,000 participants and corresponding baseline examinations by March 2021. However, delays from natural disasters, political unrest, and the COVID-19 pandemic shifted this timeline to December 2021 for completion of recruitment and baseline visits and December 2023 for completion of the 2-year follow-up. Preliminary analysis of the main study aims is conducted for every 500 enrolled participants, and final results will be reported upon enrollment of the target sample size (n = 2,000).

Safety, quality assurance, and data management

Several strategies are used to ensure consistency and integrity of protocol implementation and data collection, as well as the safety of staff and participants. First, standard operating procedures are in place for all protocols, including uniform biosample collection and processing, andquestionnaire implementation. RAs are thoroughly trained and retrained in these procedures. They follow scripted instructions for all communication with participants, including the consenting process. RAs wear laboratory coats identified with the study name, gloves, and any other protective gear needed. Participants’ confidentiality, safety, and comfort are highly safeguarded, with various checkpoints during the interview. Any minor or serious adverse event that affects the well-being of a participant or staff member or the study procedures is promptly reported and corrected. Process evaluation is conducted weekly with the whole team.

Instruments used for measurements are tested and calibrated periodically. RAs use identical portable instruments for uniformity across sites. To standardize laboratory assays, Toledo conducts all laboratory assays, quality assessment, and technical support and transfers results and extra biosamples to FDI Clinical Research securely and promptly. Assays, reagents, and equipment are monitored monthly. All freezers used for long-term storage of labeled biosamples at FDI have locks, temperature monitors, and backup generators to avoid loss of samples.

Nearly all data are entered in real-time software using handheld tablets, the exception being data collected during door-to-door enumeration, which is collected on paper forms and later entered into a database. The real-time Web-based electronic data capture tool REDCap (“Research Electronic Data Capture”) hosted at the Harvard T.H. Chan School of Public Health securely captures data and runs quality checks (50). RAs have hard (paper) copies of the questionnaires as a backup in case of power failure or loss of Internet service. Quality assurance efforts include training RAs in data collection and data entry, REDCap validation parameters, and weekly data checks for validity and completeness of data. Reports are shared with the team and monitored for corrections.

Participant engagement and follow-up

We implement multiple strategies to encourage attendance, participation, and continued interest in the study. Before a study visit, we make up to 4 attempts to schedule the participant’s visit, calling at different times of the day and on different days of the week. We communicate directly with participants about reminders and study information, using the participant’s preferred method of communication. We offer to cover or reimburse transportation expenses. Appointments are available on weekends or on 2 closely scheduled days for participants with time limitations.

During the visit, we have a standardized informed consent process in the form of a narrated video with clear and simple language and a large font for participants with visual or hearing impairments. Participants are allowed breaks and are given snacks after fasting measurements are taken. All instructions and interview content are presented at a suitable literacy level. RAs are trained in culturally sensitive interviewing (including familiarity, trust, and respect), handling sensitive information, and noninvasive measurement. We record multiple ways of contacting the participant, as well as the names and contact information of 4 people who would be able to reach the participant; this information is updated yearly. Lastly, we allow for completion of some questionnaires by phone if a participant requests to leave the visit early.

After the visit, we send a letter containing the clinical results to consenting participants for discussion with their primary-care provider, along with health-related educational materials with appropriate cultural, linguistic, and literacy content. We send birthday and holiday postcards on behalf of the study and give participants small gifts labeled with the study logo and contact information. In the case of natural disasters or situations that may disrupt residents of PR, the team sends information to our participants on how to access free or low-cost services (including mental health services) and how to stay in touch with us. Throughout the study, we reiterate our appreciation for their participation and the benefits of the study to the general PR community.

Coordinated operational strategies

PROSPECT was designed with 4 coordinated operational strategies (Figure 2). First, research productivity is at the forefront. PROSPECT is expected to fill a gap in existing knowledge on CVD risk factors in PR, as well as answer important research questions on how these risks are associated and potentially prevented. By synergizing our designs and protocols with those of other existing epidemiologic cohorts of Hispanics/Latinos (i.e., the Boston Puerto Rican Health Study and the Hispanic Community Health Study/Study of Latinos) (8, 12), we will be able to pool data from these studies for comparison and systematic reviews.

Figure 2.

Coordinated operational strategies led by the Puerto Rico Observational Study of Psychosocial, Environmental, and Chronic Disease Trends (PROSPECT).

Second, we aim to expand the research infrastructure in PR. This includes establishing a biorepository for future use in research studies, a registry of participants engaged in research, a source of skilled staff, and an islandwide network of primary-care clinics with access to medical records and the dedication and capacity to conduct clinical research. We have partnered with the Puerto Rico Vector Control Unit of the Puerto Rico Science, Technology, and Research Trust (San Juan, PR) to pilot-test a new cellphone-based application based on ArcGIS Workforce (ESRI, Redlands, California) and ArcGIS Survey123 (ESRI) that expedites assignment of enumeration tasks and displays up-to-date maps, accurate travel times, and real-time geolocation of enumerated households in a Web dashboard; its applicability in PROSPECT will enable implementation in other research studies. PROSPECT’s framework provides a unique opportunity for ancillary and future studies. To date, we are implementing 4 ancillary studies: 1) a mixed-methods study on food and water access, social support, and stress responses after natural disasters; 2) a study on seroprevalence of vectorborne diseases in partnership with the PR branch of the Centers for Disease Control and Prevention; 3) a mixed-methods project to assess food insecurity, access to food and health-care services, and glycemic control during the COVID-19 pandemic among participants with diabetes; and 4) a pilot program to improve access to healthy foods and cardiometabolic health by providing fresh produce from local farmers to participants with food insecurity. There is also a parallel study on CVD risk factors among young adults being led by the University of Massachusetts Medical School (Worcester, Massachusetts) and the University of Puerto Rico Medical Sciences Campus (San Juan, PR), the Puerto Rico Young Adults’ Stress, Contextual, Behavioral and Cardiometabolic Risk (PR-OUTLOOK) Study, which builds from and adapts PROSPECT’s framework.

Third, we support capacity-building, education, and training of established and new investigators. PROSPECT provides opportunities for career development and research experience for many investigators in multiple fields, especially underrepresented minorities. Besides offering these opportunities to our staff and external investigators, we work closely with higher education institutions in the mainland United States and PR to extend them to students and trainees.

Last, we conduct community outreach and dissemination of results in parallel with public health initiatives aimed at preventing and controlling chronic diseases in PR. For this, we have partnered with multiple professional and public health organizations, including the College of Nutritionists and Dietitians of Puerto Rico; the Society for Prevention of Cardiovascular Diseases of Puerto Rico; the Puerto Rico Science, Technology, and Research Trust; the Puerto Rico Public Health Trust; and the Puerto Rico Consortium for Clinical Investigation. Our team shares information about the study with partner clinics and the community at large. We regularly share a community report, which is a simple, brief, graphic-based pamphlet that summarizes preliminary results. As the study begins to identify priority needs for clinical and public health prevention programs and policies, we will connect with stakeholders and agents of change through an initiative supported by a Robert Wood Johnson Foundation Culture of Health Leaders Award.

DISCUSSION

PROSPECT has faced challenges, yet has attained many successes during its launch (Figure 3). We expected the household enumeration process to be strenuous and outdated, as we only had availability of 2010 US Census data, but Hurricane María deeply altered the infrastructure of the island. However, enumeration proved to be more time- and effort-consuming and has produced fewer participants than anticipated, due to the extent of missing, vacant, or unresponsive houses. Moreover, numerous unforeseen events have delayed the timeline. Having alternative plans has been key in surpassing these challenges. For example, use of community-based strategies has accelerated recruitment and suggests that this population is open to participating in health research. Our fast response to disasters has kept PROSPECT on track and has helped us gain real-time knowledge about health and behaviors in times of distress. The pilot study, PRADLAD, taught us challenges, opportunities, and lessons to implement in PROSPECT (10). One fundamental lesson is to gain the highest level of trust, respect, and full commitment to the study from participants, team members, and partners. To attain this, we have open and frequent communication, and we are genuinely attentive to the ideas and needs of all parties involved. As a result, PROSPECT has reached initial milestones with minimal delays, secured multiple ancillary projects, and gained multilevel partnerships and is actively informing policy and community health.

Figure 3.

Lessons learned from the Puerto Rico Observational Study of Psychosocial, Environmental, and Chronic Disease Trends.

Despite its groundbreaking nature, PROSPECT has some limitations. First, the sampling methodology—sampling through household enumeration—was designed to capture a probabilistic sample, but enumeration is challenging, and the community-based sampling may reduce full representation. However, the widespread community recruitment efforts increase the likelihood of selecting a general population, and we periodically check our sample’s characteristics against broader sociodemographic benchmarks to ensure representation. Second, while we designed the study similarly to other cohort studies of Hispanic/Latinos residing in the United States, it may not be feasible to directly compare all of our results with findings on other Puerto Rican populations or persons of other ethnic backgrounds. Finally, our study is limited to adults aged 30–75 years, since most chronic conditions occur at this life stage; further studies should assess CVD risk factors in other age groups.

To our knowledge, PROSPECT is the first islandwide longitudinal study designed to capture data on trends in chronic diseases and the psychosocial, lifestyle, and environmental contributors to these conditions and to inform health-care providers, public health agencies, and the general PR community on strategies for preventing disease. Novel aspects include the ability to survey the whole island, detect urban/rural variations, and assess disease after natural disasters. The comprehensive and meticulous data collection is helping establish a sustainable framework to support the study’s operational strategies. PR is at the juncture of socioeconomic, political, and environmental adversities, while simultaneously driving innovative scientific and technological advancements with homegrown talent. Our study helps address these issues and promote the local health sciences field, with much promise for broader application to other populations. Given its merits, PROSPECT has high potential for informing effective culturally tailored policies and programs designed to improve health in PR.