abstract SUPPLEMENT ARTICLE

Transcription

1 The Ntionl Children s Study: Recruitment Outcomes Using the Provider-Bsed Recruitment Approch Dniel E. Hle, MD, Shron B. Wytt, PhD, RN, CANP, b, Stephen Buk, ScD, c Debr Cherry, MD, MS, d Kendll K. Cislo, MPH, e Donld J. Dudley, MD, f Perl Ann McElfish, PhD, MBA, g Gwendolyn S. Normn, PhD, MPH, h Simone A. Reynolds, PhD, MPH, i Ann Mri Sieg-Riz, PhD, j Sndr Wdlinger, MS, RRT, k Cheryl K. Wlker, MD, l Jmes M. Robbins, PhD m OBJECTIVE: In 2009, the Ntionl Children s Study (NCS) Vngurd Study tested the fesibility of household-bsed recruitment nd prticipnt enrollment using birth-rte probbility smple. In 2010, the NCS Progrm Office lunched 3 dditionl recruitment pproches. We tested whether provider-bsed recruitment could improve recruitment outcomes compred with household-bsed recruitment. METHODS: The NCS imed to recruit 18- to 49-yer-old women who were pregnnt or t risk for becoming pregnnt who lived in designted geogrphic segments within primry smpling units, generlly counties. Using provider-bsed recruitment, 10 study centers engged providers to enroll eligible prticipnts t their prctice. Recruitment models used different levels of provider enggement (full, intermedite, informtion-only). RESULTS: The percentge of eligible women per county rnged from 1.5% to 57.3%. Across the centers, 3371 potentil prticipnts were pproched for screening, 3459 (92%) were screened nd 1479 were eligible (43%). Of those 1181 (80.0%) gve consent nd 1008 (94%) were retined until delivery. Recruited prticipnts were generlly representtive of the county popultion. CONCLUSIONS: Provider-bsed recruitment ws successful in recruiting NCS prticipnts. Chllenges included time-intensity of engging the clinicl prctices, differentil willingness of providers to prticipte, nd necessry relince on providers for prticipnt identifiction. The vst mjority of prctices cooperted to some degree. Recruitment from obstetric prctices is n effective mens of obtining representtive smple. bstrct NIH Deprtment of Peditrics, University of Texs Helth Science Center t Sn Antonio, Sn Antonio, Texs; b School of Nursing, University of Mississippi Medicl Center, University of Mississippi, Jckson, Mississippi; c Deprtment of Epidemiology, Brown University School of Public Helth, Brown University, Providence, Rhode Islnd; d Division of Generl Internl Medicine, Deprtment of Internl Medicine, University of Wshington, Settle, Wshington; e Deprtment of Epidemiology nd Biosttistics, Michign Stte University, Est Lnsing, Michign; nd f Division of Mternl nd Fetl Medicine, Deprtment of Obstetrics nd Gynecology, University of Virgini, Chrlottesville, Virgini; g Deprtment of Internl Medicine, Univesity of Arknss for Medicl Sciences Northwest, Fyetteville, Arknss; h Deprtment of Oncology, Wyne Stte University School of Medicine, Wyne Stte University, Detroit, Michign; i Deprtment of Epidemiology nd Biosttistics, School of Public Helth, Stte University of New York Downstte Medicl Center, Brooklyn, New York; j Deprtment of Epidemiology nd Nutrition, Gillings School of Globl Public Helth, University of North Crolin, Chpel Hill, North Crolin k Rymond G. Perelmn Center for Cellulr nd Moleculr Therpeutics, University of Pennsylvni School of Medicine, The Children's Hospitl of Phildelphi, Phildelphi, Pennsylvni; l Deprtment of Obstetrics nd Gynecology, University of Cliforni Dvis School of Medicine, Scrmento, Cliforni; nd m Deprtment of Peditrics, University of Arknss for Medicl Sciences, Arknss Children's Hospitl, Little Rock, Arknss Decesed Dr Hle conceptulized the mnuscript nd designed the nlysis; cquired, nlyzed, nd interpreted the dt; drfted the initil mnuscript; coordinted incorportion of the criticl review nd suggested revisions of ll uthors; Drs Wytt, Cherry, Dudley, Normn, Reynolds, Sieg-Riz, Wlker, nd Robbins, Ms Cisco, Ms McElfish, nd Ms Wdlinger cquired, nlyzed nd interpreted the dt, nd criticlly reviewed nd revised the mnuscript; Dr Buk nlyzed nd interpreted To cite: Hle DE, Wytt SB, Buk S, et l. The Ntionl Children s Study: Recruitment Outcomes Using the Provider-Bsed Recruitment Approch. Peditrics. 2016;137(s4):e E PEDIATRICS Volume 137, Number s4, June 2016 :e SUPPLEMENT ARTICLE

2 In this report, we describe the experience of the 10 study centers (SCs) tking prt in provider-bsed recruitment (PBR). The primry recruitment site for the PBR SCs ws the prentl cre provider office. We hypothesized tht, by recruiting in n environment tht ws devoted to the cre of women who were pregnnt or t risk for becoming pregnnt, recruitment rtes could be mximized for both pregnncy nd prepregnncy cohort. Enrollment through providers ws expected to increse the efficiency nd effectiveness of recruitment. Additionlly, recruitment of prticipnts in this venue llowed the study to be introduced by known nd trusted individul, the prentl cre provider, nd offered the opportunity for n NCS recruiter to hve fce-to-fce interction with potentil prticipnt in this neutrl environment. PBR hs been extensively used for pregnncy cohorts, ll of which hve inclusion criteri relted to specific disese process or tretment regimen In this novel NCS experiment, the primry inclusion criterion ws the prticipnt s ddress. This rticle describes the recruitment nd retention of NCS prticipnts using PBR. METHODS Detils regrding the selection of primry smpling units (PSUs) nd subsequent determintion of strt nd recruitment segments were provided in the introductory rticle, s were the criteri for the selection of study prticipnts. Thus, we focus on the ctivities tht were unique to the PBR pproch. Although the predominnt providers of prentl cre services in ll 10 study loctions were obstetricins, there were other providers who included fmily physicins, nurse prctitioners, nd midwives. To encompss this vribility, ll individuls providing prentl cre services re subsequently described s providers. We developed severl strtegies to engge the providers in ech PSU. The locl NCS study tem first identified which entities provided prentl cre services to women in the PSU. In most counties, prentl cre providers tended to cluster ner birthing fcilities. In some counties there were identifible clusters of birthing hospitls s well, wheres in others, birthing hospitls were disseminted. In the high popultion counties with mny choices for birthing hospitls, we could not discern cler pttern of pregnnt women preferring the prentl cre provider or birthing hospitl tht ws most proximte to their homes. Thus, ll provider loctions serving women in the PSU were considered potentil recruitment venues. Becuse locl investigtors were leding ech recruitment effort, mny of the providers were lredy known nd, in mny cses, hd collborted on other projects. This ws especilly true in the rurl counties. Other pproches used to ensure tht ll providers were identified included review of county birth certifictes, serching delivery hospitl Web sites, obtining lists of hospitl delivery privileges, exploring dvertisements for obstetricl services, using miling lists for obstetricl grnd rounds, online serches, nd questioning nursing stff in lbor nd delivery units. Once prticulr prctice ws identified, dditionl detils bout the prctice, such s office loctions nd hospitl ffilitions, were sought from ll sources. This dditionl informtion permitted the prctice recruitment tem to be knowledgeble bout the prctice before ctul contct with the provider office occurred. Systemtic efforts were mde to inform providers bout the NCS before engging them directly. Presenttions regrding the gols nd strtegies of the NCS were mde t grnd rounds, deprtmentl events, nd t locl, regionl, nd ntionl meetings. Articles feturing the NCS were published in professionl society newsletters nd in locl newsppers. Fetures bout the NCS were ired on locl rdio nd television. Prctices were typiclly contcted by locl NCS investigtor or senior stff member. In mny circumstnces, the locl NCS investigtor provided clinicl consulttive services for the providers. Also, providers hd often been trined t the cdemic centers nd hd estblished reltionships with NCS investigtors. Often, office mngers nd nursing personnel in the providers offices served s the primry point of contct. They proved to be instrumentl in llowing ccess to ptient ddresses nd clinicl schedules. These initil contcts were followed by further interctions to estblish the logistics for dy-to-dy opertions t ech prctice. The gols of these interctions were fivefold: (1) to educte the stff bout the NCS, (2) to obtin permission to recruit potentil prticipnts, (3) to determine process for confirming ptient ddress eligibility, (4) to negotite the exct recruitment pproch within the individul office, nd (5) to build trust of NCS personnel with office stff. Our guiding principles for involvement in provider offices were (1) to customize the recruitment process bsed on provider recommendtions, (2) to minimize study-relted burden for the clinic, nd (3) to minimize interruption in clinic flow. In dvnce of recruitment initition, the customized implementtion pln for NCS-relted ctivities ws reviewed with the stff t ech loction. The pproch to ctul prticipnt recruitment in the provider s office ws guided by provider preference, nd fell into 3 brod ctegories of enggement. The most limited S240 HALE et l

3 pproch ws to plce brochures nd posters contining informtion for prticipnt self-referrl in the provider office. No mention of the NCS ws mde to the potentil prticipnt by the provider or stff unless the ptient sked. The intermedite pproch involved study cknowledgment by the provider. There were severl vrints of this ctivity, rnging from hnding potentil prticipnt brochure t registrtion, to miling n introductory letter on the provider s sttionery before clinic visit, to mentioning the NCS during the provider encounter. The most ctive pproch ws endorsement of the NCS by the provider, ccompnied by introduction of potentil prticipnt to n NCS recruiter in the office. For loctions functioning in 1 of the ltter 2 ctegories, studyrelted mterils were plced in the chrts of potentil prticipnts in dvnce of the visit. For the offices tht ctively endorsed the study, NCS recruitment stff coordinted with the loction to hve recruiter on-site t the time of the potentil prticipnt s ppointment. When possible, finl eligibility screening for ge nd pregnncy sttus occurred in the office. The collbortive greements with the prctices rnged from simple verbl greement to forml contrct. As much s possible, the NCS investigtors deferred to the preference of the prctice. Becuse prctices were only providing informtion bout the NCS to their ptients, providers were not considered to be engged in reserch by the federl Office of Humn Reserch Protections nd most locl institutionl review bords (IRBs). As consequence, most privte prctices did not require the locl NCS tem to obtin dditionl IRB pprovls. Helth Informtion Portbility nd Accountbility Act (HIPAA) wivers were necessry to permit NCS stff to review ptient informtion. Hospitl-bsed nd university-ffilited clinics required study pprovl by their IRB of record. Providers were consistently informed tht NCS stff hd undergone bckground checks, HIPAA nd IRB trining, nd pproprite helth screening nd vccintions. Documenttion ws provided when requested. NCS stff members working in institutionl nd hospitlbsed clinics were required to complete fcility-mndted reserch ethics trining. Activities requiring n interfce with provider offices included verifying tht the womn lived in the selected segment, introducing the NCS to potentil prticipnts, nd ctul recruiting done by NCS stff. Although ech office ws unique, ddress-eligibility review ws generlly done in 1 of 3 wys: (1) ddress lists of upcoming prentl ppointments were provided to NCS stff, typiclly vi secure fx or uploded to secure server; (2) provider stff prescreened upcoming prentl visits, by using study-provided ddress tool or nother ssistive prmeter (eg, zip code known to include secondry smpling unit) nd then provided this informtion to NCS stff; or (3) NCS stff ws given limited ccess to ptient ppointment schedules, which were regulrly reviewed by NCS stff. The provider office ws informed when ddresseligible women were identified nd n pproprite pln for contct ws developed. Recruiters were permitted considerble flexibility in obtining consent, llowing them to be responsive to potentil prticipnt preferences, such s the inclusion of spouse in the discussion or meeting the individul t different time to obtin consent. A vriety of other ctivities were undertken by individul sites in n effort to boost recruitment. These ctivities were neither mndted nor prohibited by the NCS protocol. The choice of ctivities ws bsed on locl preferences, experiences, opportunities, nd resources nd ws significntly ffected by the popultion density nd demogrphics of the PSU. Advntge ws tken of the public reltions expertise of the locl institution chrged with recruitment in prticulr PSU. Recruitment ctivities included press releses, ppernces of NCS personnel on locl news progrms, humn interest stories in community newsppers, nd stff ppernces t helth firs. Some PBR SCs undertook pid dvertising cmpigns, wheres others used tilored mss miling within the designted segments. Other SCs used billbords on mjor streets ner their segments or prentl cre provider offices. Some NCS stff members took prt in tours provided for pregnnt women by birthing hospitls. This ws useful wy for NCS stff to become fmilir with the hospitl s well s good opportunity to present the NCS to pregnnt women. In severl counties, pregnncy screening centers were menble to llowing NCS recruitment within their fcility. Some sites sttioned NCS recruiters t locl Specil Supplementl Nutrition Progrm for Women, Infnts nd Children nd prentl support services sites, or prticipted in childbirth nd fthering clsses. Community dvisory bords (CABs) were estblished erly in the development of ech NCS SC to ssist with community enggement, select study segments, nd guide locl implementtion. In the rurl counties, the CAB ws composed of individuls who would be knowledgeble bout the vrious popultions nd issues in the PSU, nd included elected officils; known community ctivists; business, church, nd school leders; nd locl medicl cre providers. Becuse of the greter diversity within the urbn counties, nd smller percentge of the popultion tht ws eligible PEDIATRICS Volume 137, number s4, June 2016 S241

4 TABLE 1 County Demogrphics nd Orgniztion of Mternity Cre Totl Prentl Totl Prctice Averge No. of Address Cre Providers b Loctions c Eligible Pregnnt Women per Loction c Totl Women Ages Totl County Approximte % of Yers Births Pregnnt Women Address Eligible County Nme Approximte Density, Popultion Persons per Squre Mile Lmr, TX Schuykill, PA Benton, AR Hinds, MS All rurl Durhm, NC Bexr, TX New Hven, CT Scrmento, CA Providence, RI Wyne, MI All urbn , popultion density is county specific US Census Popultion Estimtes. b This includes ll providers who delivered n infnt to county resident. In some counties, most infnts were delivered by only few providers. c This includes ll loctions where prentl visit occurred. In some counties, most prentl ptients were seen t limited number of loctions. to prticipte, the CAB ws more typiclly structured to ssist the NCS SC tem in the identifiction of individuls specific to secondry smpling unit who could provide guidnce bout recruiting in prticulr segment. Additionlly, mny CABs included providers who could dvise SCs regrding potentil strtegies to engge other providers in the community. Once the PBR strtegy ws nnounced, severl CABs were restructured to reflect the ltered recruitment strtegy, specificlly by incresing the number of providers, office dministrtors, hospitl dministrtors, nd lbor nd delivery nurses. All dt collected from the SCs were trnsmitted to centrl dt repository nd formed the bsis of the outcomes reported in this study. RESULTS The popultion nd popultion density vried cross the 10 PBR PSUs (Tble 1). As consequence, the geogrphic size of the selected segments rnged from few city blocks to mny squre miles. The popultion density nd totl popultion in the re ffected the number of prentl cre providers nd birthing fcilities. These fctors presented unique logisticl chllenges for project implementtion. We defined county-wide popultion density of <500 persons per squre mile s rurl nd density 500 persons per squre mile s urbn. As shown in Tble 1, the popultion density of the PBR counties rnged from 55 to 2945 persons per squre mile. Thus, 4 PBR counties were rurl nd 6 were urbn. The totl number of women between the ges of 18 nd 49 yers nd the number of births ech yer reflected the totl popultion of the county. The fertility rte (births to women, ges 18 49) rnged from 5.45% to 7.14% in individul counties, but ws similr S242 HALE et l

5 TABLE 2 Opertionl Chrcteristics of Prctice-Bsed Recruiting Stffing Model c No. of Prctice Prctice Loctions Engged, n (%) County Time to Strt, wk Durtion of Recruitment, wk b Loctions d Lmr, TX Tsk cores/subcontrct 9 9 (100) Schuykill, PA Tsk cores 5 5 (100) Benton, AR Cse mngement (100) Hinds, MS Tsk cores/ subcontrct (100) All rurl 52.0 ± ± (100) Durhm, NC Tsk cores/subcontrct (100) Bexr, TX Cse mngement (48) d New Hven, CT Cse mngement (68) d Scrmento, CA Cse mngement (64) d Providence, RI Tsk cores (47) d Wyne, MI Tsk cores (45) d All urbn 54.7 ± ± (55) Strt ws the dte on which the first prticipnt could hve been recruited. Averge time from contrct wrd to strt ws 54 ± 4 weeks. b Recruitment ws ended before ll offices were recruited. c Tsk cores hd personnel ssigned to specific ctivities (eg, recruitment, dt collection). Some tsks were subcontrcted. Cse mngement hd n identified individul who lwys ws the point of contct for specific prticipnt. d Prctice loctions were used becuse ech office hd to be individully pproched nd mnged. when compring the rurl (6.78%) nd urbn (6.60%) counties. As consequence of the wide popultion size vribility mong the 10 PBR counties, the percentge of ddresseligible women vried from 57.3% in rurl Lmr County, Texs, to 1.5% in urbn Bexr County, Texs. Overll, in rurl counties, 1 of every 6 women ws eligible, wheres in urbn counties, 1 of every 41 women ws ddress eligible. There ws considerble diversity in provider orgniztionl structure, rnging from solo prctitioner to lrge group prctices with mny providers nd offices. Prctice orgniztion tended to be more complex in high birth counties. Becuse most women received their prentl cre t specific loction, we chose to consider ech prentl cre office loction s unique. Women receiving obstetricl cre outside of the PSU rnged from <1% to 18% nd tended to be higher in counties with few providers. The verge nnul number of ddress-eligible women per prctice loction vried from 3 to 77. More ddress-eligible women per loction were noted in the rurl counties (men = 23) thn in urbn counties (men = 7). Becuse the lternte recruitment strtegies, such s PBR, were intended to inform lter study ( Min NCS Study ), opertionl prmeters were trcked. Tble 2 summrizes mjor opertionl chrcteristics. The time to initite enggement of providers verged 1 yer (53 ± 4 weeks), much of which ws due to Federl Informtion Security Mngement Act 15 requirements nd lso obtining IRB pprovl. There ws little difference between rurl nd urbn sites for time to strt (52.0 vs 54.7 weeks) or the recruitment period (47.0 vs 44.3 weeks). As noted in Tble 2, stffing models vried by site. There were 2 generl models. The tsk core pproch ssigned different personnel to specific tsks, such s recruitment or dt collection. The cse mngement model ssigned specific stff member to ll ctivities involving ech recruited individul. Becuse recruitment ws stopped premturely, we were unble to determine which one model ws superior. By design, pproximtely equl numbers of ddress-eligible women lived in ech county. There ws vrible success in screening nd consenting potentil prticipnts (Tble 3). The 4 rurl SCs verged slightly more women screened (370 per county) thn the 6 urbn SCs (327 per county); however, there ws considerble vribility within both groups. The percentge of women confirmed to be study eligible by ddress in the rurl counties ws lmost twice tht in the urbn counties (59% vs 31%), lthough gin there ws substntil vribility within the 2 groups. Consent rtes were somewht higher in the urbn counties. Across the 10 counties, the overll consent rte ws 80%, higher thn the predicted consent rte of 65%, nd well bove tht reported in the initil Vngurd Study recruitment effort (61%). In fct, only 1 of the SCs hd consent rte less thn the predicted consent rte. Success t mintining prticiption in the study from consent until birth ws 94% overll with little difference between the rurl nd urbn counties (92% vs 96%). Four of the 10 counties (Hinds, Durhm, Wyne, nd Bexr) merit comment becuse most consented women were minority women, lrgely of Hispnic ethnicity. The overll retention rte in these counties ws similr to those in the other counties (95% vs 93%). The lowest rte of retention ws in rurl Hinds County (86%), wheres the highest rte of retention ws in urbn Bexr County (98%), both PEDIATRICS Volume 137, number s4, June 2016 S243

6 TABLE 3 Prticipnt Identifiction, Recruitment, nd Retention Percent of Pregnnt Consented Women Retined t Birth No. of Pregnnt Women Retined From Consent to Birth No. of Consented Women Who Are or Becme Pregnnt Consent Rte, % No. of Women Consented Eligibility Rte, % No. of Women Determined to Be Study Eligible Screening Completion Rte, % No. of Women Screened County Nme No. of Women Approched for Screening Lmr, TX Schuylkill, PA Benton, AR Hinds, MS All rurl Durhm, NC Bexr, TX New Hven, CT Scrmento, CA Providence, RI Wyne, MI All urbn Dt points for Lmr, TX, re bsed on the study center cse logs nd not on the dt trnsmitted to the centrl dt repository. counties with minority popultions. Overll refusl rtes, defined s study withdrwls or loss-of-contct, were low. Insted, refusls were typiclly pssive, such s not responding to phone clls or keeping ppointments. Tble 4 compres chrcteristics of enrolled prticipnts to ll women giving birth in the county. Some cution is required in interpreting these comprisons becuse recruitment ended before ll of the providers in the urbn counties hd been pproched to prticipte in the NCS. As shown in Tble 4, women from the 2 lrgest minority groups, Hispnic nd non-hispnic blck, were enrolled in pproximtely the sme proportion s ll women giving birth in the county. In generl, women with nnul household incomes <$ were represented in t lest equl proportions mong consented prticipnts. Nine of the counties enrolled higher percentge of women with eduction greter thn high school. For some counties, the finl cohort demogrphics reflected the intentionl erly enggement of prctices serving the highest number of ddress-eligible women, leding to some skewing of the dt. This skewing would likely hve disppered once ll of the prctices becme engged. In 6 counties, 10% or more of consented women reported primry lnguge other thn English. In generl, non English-speking prticipnts were menble to consenting to the study, s reflected in the dt in Tble 4. DISCUSSION The PBR method offered unique chllenge becuse before prticipnt recruitment begn, provider hd to gree to llow ccess to ptients, nd ccess to HIPAAprotected informtion hd to be obtined to determine ddress eligibility. Although initilly envisioned s mjor chllenge, given the nticipted complexity S244 HALE et l

7 TABLE 4 Comprison of NCS Pregnnt Enrolled Women to County Reference Popultion Annul Income <$50 K, % Primry Lnguge Other Thn English, % County Non-Hispnic White, % Hispnic, % Non-Hispnic Blck, % High School Eduction or Less, % Ntlity NCS Ntlity NCS Ntlity NCS Ntlity NCS ACS NCS NCS Lmr, TX Benton, AR Schuykill, PA Hinds, MS b Durhm, NC b Bexr, TX b New Hven, CT Scrmento, CA Providence, RI Wyne, MI b ntlity dt. The 3 mjor ethnic groups re shown for ech county. Other ethnic nd rcil groups ccounted for 0% to 18.1% of the totl popultion in vrious counties., The number of individuls in these cells is lower thn the NCS disclosure voidnce threshold or could be derived from informtion provided in other cells. Americn Community Survey (ACS) benchmrk dt on fmily income for girls nd women ge 15 to 50 with birth in pst 12 months ( yer estimtes). b Counties in which most births were to minority women. nd longevity of the NCS, provider enggement proved reltively strightforwrd. The min brrier to prctice enggement ws the time required for the NCS SC tem to estblish contct nd subsequently to customize study-relted ctivities to the prctice. This is reflected in the fct tht SCs with >50 prctice loctions hd not fully pproched ll of the prctices to prticipte t the time tht prticipnt recruitment ws hlted. Despite the chllenge of engging multiple prctice loctions, the PBR group demonstrted tht using provider-bsed strtegy ws fesible in urbn nd rurl settings. This pproch ws cceptble to prticipnts, with excellent rtes of retention nd miniml ctive withdrwls. An dded element of fesibility for the PBR pproch ws whether diverse providers, prctices, nd loctions would gree to tke prt. In the rurl counties, there ws consistent support for the NCS: lmost ll of the prctices, regrdless of institutionl ffilition or orgniztionl structure, greed to prticipte, lthough some required compenstion. We did not determine if compenstion to prctices ws criticl in their decision to prticipte in the study, or if more widespred use of finncil incentives would hve improved provider prticiption rtes. In the urbn counties, prctices were still being contcted t the time tht prticipnt recruitment ws hlted; however, in most counties, significnt mjority of prctices greed to prticipte. Office-bsed recruitment ws very cceptble to the providers nd there ws only 1 provider withdrwl cross lmost 300 office loctions. A key NCS gol ws to obtin ntionlly representtive smple reflective of the diverse demogrphics nd environmentl exposures of pregnnt women. During the reltively short recruitment period, cross the 10 PEDIATRICS Volume 137, number s4, June 2016 S245

8 PBR counties, 1181 consented, 1074 becme pregnnt, nd 1008 were retined through delivery. Our results suggest tht sufficiently representtive smple cn be ccomplished by using PBR. Some of the vribility of recruitment success with specific rcil, ethnic, lnguge, eductionl, or economic groups likely reflects the strtegies used by ech SC to enroll prctices, especilly in the more populous counties, where not ll prctices hd been engged t the time of cesstion of prticipnt recruitment. Weknesses of our study include limited time for recruitment of providers, prticulrly in the urbn counties, nd incomplete dt collection on the success of the different pproches (full, intermedite, informtion-only) deployed in provider offices. Strengths of the PBR included the diverse methodologies used to ensure tht provider s concerns were ddressed nd their prticiption s recruitment sites for the NCS ws secured. Additionlly, we were ble to compre our dt with county birth dt to determine tht we were ble to recruit reltively representtive smple of reproductive-ged women. CONCLUSIONS We estblished tht PBR ws successful in enrolling generlly representtive smple of reproductive-ged women in defined geogrphic region. Although efforts t recruiting community-bsed providers to prticipte in reserch trils ws time-consuming nd lbor intensive, we found tht, with cretive pproches to ensure tht prctice ctivities were minimlly ffected, providers were generlly supportive of prticiption in the NCS. Our successful model for enrolling representtive smple by using PBR should inform future studies of key obstetric nd neontl outcomes. ACKNOWLEDGMENTS We thnk the mny providers in our communities who opened their offices to this project. Without their ssistnce, this recruitment pproch would not hve been possible. We thnk the fmilies nd community members who prticipted in, or contributed to, the NCS providerbsed project. We cknowledge the mny stff members t ech of the sites who worked tirelessly on this project. ABBREVIATIONS CAB: community dvisory bord HIPAA: Helth Informtion Portbility nd Accountbility Act IRB: institutionl review bord NCS: Ntionl Children s Study PBR: provider-bsed recruitment PSU: primry smpling unit SC: study center the dt, nd criticlly reviewed nd revised the mnuscript; nd ll uthors pproved the finl mnuscript s submitted nd greed to be ccountble for ll spects of the work. This tril hs been registered t www. clinicltrils. gov (identifier NCT ). DOI: /peds E Accepted for publiction Mr 1, 2016 Address correspondence to Dniel E. Hle, MD, Deprtment of Peditrics, University of Texs Helth Science Center t Sn Antonio, 7703 Floyd Curl Dr, Sn Antonio, TX E-mil: hle@uthscs.edu PEDIATRICS (ISSN Numbers: Print, ; Online, ). Copyright 2016 by the Americn Acdemy of Peditrics FINANCIAL DISCLOSURE: The uthors hve indicted they hve no finncil reltionships relevnt to this rticle to disclose. FUNDING: Supported in prt by Ntionl Institute for Child Helth nd Humn Development contrcts HSN C, HHSN C, HHSN C, HHSN C, HHSN C, HHSN C, HHSN C, HHSN C, HHSN C, nd HHSN C. Funded by the Ntionl Institutes of Helth (NIH). POTENTIAL CONFLICT OF INTEREST: The uthors hve indicted they hve no potentil conflicts of interest to disclose. REFERENCES 1. The Children's Helth Act of Pub.L , 114 Stt (Oct. 17, 2000) 2. Svitz DA, Zeger SL, Golding J, et l. Finl Report from the Ntionl Children's Study Smpling Design Workshop. My 9, Wshington DC: Ntionl Reserch Council, Institute of Medicine, Ntionl Acdemies Press; Hirschfeld S, Songco D, Krmer BS, Guttmcher AE. Ntionl Children's Study: updte in Mt Sini J Med. 2011;78(1): Trsnde L, Andrews HF, Gornson C, et l. Erly experiences nd predictors of recruitment success for the Ntionl Children s Study. Peditrics. 2011;127(2): Kerver JM, Elliott MR, Normn GS, et l; MANCS Executive Committee. Pregnncy recruitment for popultion reserch: the Ntionl Children s Study vngurd experience in Wyne County, S246 HALE et l

9 Michign. Peditr Perint Epidemiol. 2013;27(3): Ntionl Reserch Council. The Ntionl Children s Study Reserch Pln: A Review. Wshington, DC: The Ntionl Acdemies Press; Montquil JM, Brick JM, Curtin LR. Sttisticl nd prcticl issues in the design of ntionl probbility smple of births for the Vngurd Study of the Ntionl Children s Study. Stt Med. 2010;29(13): Guttmcher AE, Hirschfeld S, Collins FS. The Ntionl Children s Study proposed pln. N Engl J Med. 2013;369(20): Institute of Medicine nd Ntionl Reserch Council. The Ntionl Children s Study 2014: An Assessment. Wshington, DC: The Ntionl Acdemics Press; Jddoe VW, Mckenbch JP, Moll HA, et l. The Genertion R Study: design nd cohort profile. Eur J Epidemiol. 2006;21(6): Mgnus P, Irgens LM, Hug K, Nystd W, Skjerven R, Stoltenberg C; MoB Study Group. Cohort profile: the Norwegin Mother nd Child Cohort Study (MoB). Int J Epidemiol. 2006;35(5): Eskenzi B, Gldstone EA, Berkowitz GS, et l. Methodologic nd logistic issues in conducting longitudinl birth cohort studies: lessons lerned from the Centers for Children s Environmentl Helth nd Disese Prevention Reserch. Environ Helth Perspect. 2005;113(10): Buck GM, Lynch CD, Stnford JB, et l. Prospective pregnncy study designs for ssessing reproductive nd developmentl toxicnts. Environ Helth Perspect. 2004;112(1): Buck Louis GM, Schistermn EF, Sweeney AM, et l. Designing prospective cohort studies for ssessing reproductive nd developmentl toxicity during sensitive windows of humn reproduction nd development the LIFE Study. Peditr Perint Epidemiol. 2011;25(5): Federl Informtion Security Mngement Act (FISMA) of 2002, 44 U.S.C PEDIATRICS Volume 137, number s4, June 2016 S247

10 The Ntionl Children's Study: Recruitment Outcomes Using the Provider-Bsed Recruitment Approch Dniel E. Hle, Shron B. Wytt, Stephen Buk, Debr Cherry, Kendll K. Cislo, Donld J. Dudley, Perl Ann McElfish, Gwendolyn S. Normn, Simone A. Reynolds, Ann Mri Sieg-Riz, Sndr Wdlinger, Cheryl K. Wlker nd Jmes M. Robbins Peditrics 2016;137;S239 DOI: /peds E Updted Informtion & Services References Permissions & Licensing Reprints including high resolution figures, cn be found t: This rticle cites 9 rticles, 1 of which you cn ccess for free t: #BIBL Informtion bout reproducing this rticle in prts (figures, tbles) or in its entirety cn be found online t: Informtion bout ordering reprints cn be found online:

11 The Ntionl Children's Study: Recruitment Outcomes Using the Provider-Bsed Recruitment Approch Dniel E. Hle, Shron B. Wytt, Stephen Buk, Debr Cherry, Kendll K. Cislo, Donld J. Dudley, Perl Ann McElfish, Gwendolyn S. Normn, Simone A. Reynolds, Ann Mri Sieg-Riz, Sndr Wdlinger, Cheryl K. Wlker nd Jmes M. Robbins Peditrics 2016;137;S239 DOI: /peds E The online version of this rticle, long with updted informtion nd services, is locted on the World Wide Web t: Peditrics is the officil journl of the Americn Acdemy of Peditrics. A monthly publiction, it hs been published continuously since Peditrics is owned, published, nd trdemrked by the Americn Acdemy of Peditrics, 141 Northwest Point Boulevrd, Elk Grove Villge, Illinois, Copyright 2016 by the Americn Acdemy of Peditrics. All rights reserved. Print ISSN: