Table 2. Summary of Review Articles.

Period and Methods	Inclusion and Exclusion Criteria N = Number of Studies	Intervention	Impact, Process, and Outcomes Assessed	Notes/Comments Re: QOL in Minority and Disadvantaged Populations
Eakin EG, Bull SS, Glasgow RE, Mason M. Reaching those most in need: a review of diabetes self-management interventions in disadvantaged populations. Diabetes Metab Res Rev. 2002;18:26-35.
1987-2001: MEDLINE  supplemented by review  of bibliographies from  identified studies and  reviews	Included English-language  RCTs or quasi-  experimental studies with  comparison group; DSME  intervention delivered to  an underserved/  disadvantaged group or  community.	Component DSME, group and individual  sessions, informational  mailings, media  campaign, peer support,  home visits, unstructured  sessions, videotapes,  handouts, monthly calls  from nurse	RE-AIM (Reach, Efficacy,  Adoption, Implementation,  and Maintenance)  framework used to compare and evaluate studies Physiological GHb, cholesterol, BP, weight, FBG, BMI, body composition, glucose tolerance and symptoms	Studies were conducted with  explicit focus on minority  and disadvantaged  populations.
				A summary methodologic  rating of 0 to 9 was  calculated. Scores ranged  from 2 to 8, with 4 studies  receiving a score of 7 or  more.
	Excluded descriptive reports  and studies without  comparison group.  N = 10 studies	Content  DM knowledge, early  detection, exercise,  nutrition, lifestyles,  history and culture (Pima  Indian community),  compliance, access to  primary care  Setting  Delivered in community  centers, community,  hospital clinic, and by  telephone	Behavioral  Diet, physical activity,  smoking, EtOH use  Knowledge  DM self-management  (nutrition, general DM),  changes in medication  regime  Other  Cost to provide intervention  Psychosocial  Depression, social support,  self-efficacy, and QOL  QOL: 3 of 10 studies	Authors emphasize that  DSME interventions for  underserved populations  should explicitly address  social-contextual issues.
				Authors suggest that  interventions designed to  be proactive, such as  telephone follow-up and  behaviorally focused DSME  interventions incorporated  within the primary care  visits, have successful  levels of implementation.
				QOL instruments:  Daniel et al, 1999-not  reported; Glasgow et al,  1992-assessed diabetes-  specific QOL with a  modified DCCT DQOL  scale, no between-group  statistical significance  observed; Weinberger et  al, 1995-measured HRQOL  using the SF-36; statistical  significance not observed.
Ellis SE, Speroff T, Dittus RS, Brown A, Pichert JW, Elasy TA. Diabetes patient education: a meta-analysis and meta-regression. Patient Educ Couns. 2004;52:97-105.
1990-12/2000: MEDLINE,  CINAHL, HealthSTAR,  ERIC, Science Citation  Index, PsyclNFO, CRISP,  and AADE database	Included English-language  RCTs using educational,  including nonpharmacological,  intervention intended to  improve patients’ health  status (physical,  intellectual, and/or  psychosocial),  interventions for adults,  and reporting pre- and  postintervention A1C  values (at least 12 weeks  postintervention).  N = 28 studies	Elasy’s taxonomy used to  categorize educational  interventions and assess  relationship between  specific variables  within the  interventions and  metabolic control.	Physiological  A1C  Psychosocial  HRQOL/QOL: not reported	Authors report only A1C  measures, noting other  important outcomes were  neither uniformly available  nor uniformly measured in  the literature.
		Component  Didactic, negotiated goal  setting; goal setting;  situational problem  solving; cognitive  reframing; and other  unspecified		Authors used meta-regression  to identify “components” of  the educational intervention  that best explained  variance in metabolic  control. Meta-analysis  supports the notion that  patient education improves  glycemic control.
		Content  Diet, exercise, self-  monitoring blood glucose,  basic diabetes  knowledge, medication  adherence, psychosocial,  and other unspecified		The authors examined the  impact of the number of  “episodes” as well as the  duration, and neither of  these dose-related  indicators predicted (was  an indicator) an  intervention’s success or  failure.
		Setting  No reported setting.  Referred solely to the  number of recipients		Authors note that work would  have been enhanced with  the inclusion of behavioral  outcomes.
Gary TL, Genkinger JM, Guallar E, Peyrot M, Brancati FL. Meta-analysis of randomized educational and behavioral interventions in type 2 diabetes. Diabetes Educ. 2003;29:488-501.
1966-1999: MEDLINE;  Cochrane Collaboration  database (1990-1999);  references from experts,  colleagues, previous  meta-analyses, and  review articles	Included published trials  randomized by clinician  and/or patient, sample  size ≥ 10, English  language; educational,  counseling, or behavioral  interventions aimed at  long-term self-care  behavior.	Components  Individual and group  counseling, instruction  packets and audiovisual  materials, telephone  outreach; clinician  prompting, clinician  education, computer  programs	Physiological  FBG, total glycohemoglobin,  hemoglobin A1 or  hemoglobin A1C, body  weight or BMI, blood  pressure, and lipids  Psychosocial  QOL: not reported	Authors assigned and  categorized methodologic  quality scores. Scores  were grouped into low  (<0.65), moderate (0.65-  0.79), and high (≥ 0.80),  with 5 studies receiving a  quality rating of high.
	Excluded published abstracts,  type 1, drug interventions,  and studies evaluating  short-term effects.  N = 18 studies	Content  Patient-diet, exercise,  medication regime  changes or adherence,  BGSM, and foot care		Authors found that compared  with controls, most  intervention groups  produced a decline in  glycohemoglobin.  Furthermore, group and  individual counseling  produced similar effects.
		Providers-methods to  increase patient  involvement, DM  pathophysiology,  complications, DM  education, urine testing,  and treatment/regimen  adherence  Setting  Outpatient clinic (96%)		Authors note that very few  studies reported including  African American or  Hispanic individuals.
				Authors cite few studies that  have evaluated culturally  sensitive interventions for  African Americans and  other ethnic minority  populations, an issue that  should be addressed.
				Authors note lack of QOL  measurement as a  limitation.
Glazier RH, Bajcar J, Kennie NR, Willson K. A systematic review of interventions to improve diabetes care in socially disadvantaged populations. Diabetes Care. 2006;29: 1675-1688.
1986-2004: MEDLINE,  EMBASE, CINAHL, Health  STAR, Cochrane Library,  Sociological Abstracts,  Social Science Citation  Index, and International  Pharmaceutical Abstracts	Included RCT, CT, and  before-and-after studies  with a control group;  studies aimed toward low  SES or ethnic/racial  minority or socially  disadvantaged adults  with type 1 and 2 DM.  Included studies of any  language that measured  self-management,  provider management, or  clinical outcomes.	Elasy’s intervention  taxonomy used to  describe scope and  components of the  interventions.	Physiologic  FBG.A1C, BP, BMI, lipids,  mortality, DM  complications  Patient behavior  Glucose monitoring, diet and  exercise, medication  adherence and self-  adjustment, scheduling  and/or attending  scheduled medical  appointments  Provider behavior  Management: diagnostic  testing, prescribing,  referrals, educational and  behavioral counseling  Psychosocial  QOL: not reported	Studies focused on those  groups with low SES or  belonging to an ethno-  racial minority. To evaluate  methodological quality, the  authors used an evaluation  method to identify specific  intervention features that  are associated with  successful or unsuccessful  outcomes: (1) target  of intervention,  (2) intervention design,  (3) setting, (4) delivery, and  (5) intensity and duration.
		Components  Individualized assessment  and goal setting,  individual educational  counseling, DSME,  reminder cards,  videotapes, support  groups, case management,  evidenced-based  guidelines, community  health workers/peer  educators, group visits,  treatment algorithms
		Excluded: age-specific,  gestational DM studies;  studies that only reported  hospital process of care  measures; and those not  specifying which socially  disadvantaged group.  N = 11 studies
		Content  DM knowledge, self-care,  symptoms and treatment,  exercise, nutrition, foot  care, lifestyle, monitoring;  clinician-focused  evaluation and  management of glycemic  control, associated  comorbidities and  complications		Authors suggest that cultural  tailoring, use of community  educators or lay educators,  one-on-one interventions,  individualized  assessment/reassessment,  use of treatment  algorithms, behavioral  interventions, and patient  feedback are all consistent  with positive rate  differences found in  multiple studies.
		Setting  Primary care practice sites,  hospitals, community-  based clinics, community  facilities (centers,  churches), and telephone		High-intensity (>10 contacts)  and delivery over a long  time frame (≥ 6 months)  were also associated with  success intervention in  disadvantaged groups.
Loveman E, Royle P, Waugh N. Specialist nurses in diabetes mellitus. Cochrane Database Syst Rev. 2003;(2):CD003286.
1966-11/2002: MEDLINE,  Cochrane library,  EMBASE, CINAHL, British  Nursing Index, Royal  College of Nursing Index,  HealthSTAR, BIOSIS,  PsyclNFO, Science  Citation Index, Social  Sciences Citation Index;  hand-searched relevant  journals and conference  proceedings (1990-2001),  reference lists, National  Research register, Early  Warning system and  Current Controlled Trials  registries	Included randomized  controlled and controlled  clinical trials designed to  test effects of DNS/NCM  interventions, trial  duration 6 months.	Components  DNS plus routine care  versus routine care,  automated calls with  structured messages,  treatment regime  alterations led by  DNS/NCM, treatment  recommendations to PCP  by DNS/NCM, and care  coordination within  primary care system  Content  Goal setting and self-care  Setting  Hospital, outpatient clinic,  primary care system, and  community	Physiological  All trials used A1C as  endpoint; short-term  complications  (hypoglycemic episodes,  hyperglycemic incidents)	Methodologic quality of  studies was assessed  using factors included in  Schulz and Jadad’s quality  criteria: minimization of  selection, attrition, and  detection bias. Studies  were then subdivided in  the following categories:  low, moderate, and high  risk of bias.
	Excluded studies with no  control group, group  education, no  prespecified outcomes,  other team members  other than DNS/NCM  involved delivering  intervention, nurses not  able to adjust treatments.  N = 6 studies		Other  Emergency room visits,  hospitalizations  Psychosocial  Quality of life  QOL: 1 of 6 studies with no  data presented
			Other  Prespecified outcomes  established for this  review but not included  and/or reported in trials:  long-term complications  (retinopathy, neuropathy,  and nephropathy),  mortality, BMI, costs, and  adverse effects	Authors noted that the quality  of trials was mostly not  good, making it difficult to  assess the implications for  practice. Future research  suggestions include  international observational  studies to identify roles  and time allocations;  qualitative studies and  RCTs of specialist nurse  intervention.
Norris SL, Engelgau MM, Venkat Narayan KM. Effectiveness of self-management training in type 2 diabetes: a systematic review of randomized controlled trials. Diabetes Care. 2001;24:561-587.
1/1980-12/1999: MEDLINE,  ERIC, CINAHL, restricted  to RCTs, hand-searched  several relevant journals;  abstracts and  dissertations were  excluded	Included English-language  RCTs examining the  effectiveness of self-  management training, all  or most subjects with  type 2 DM, age >18,  multicomponent  interventions included (if  educational component  evaluated separately).  Excluded children.  N = 72 studies	Components  Chart reminders, group  and individual sessions,  written information,  videos, didactic  education, individual  sessions based on  patient’s priorities, home  visits, computer  knowledge assessment  program, feedback,  behavior modification,  contracts,  demonstrations,  food logs, nutrition  goals, interactive  computer, culturally  appropriate flashcards,  pharmacist, nursing  students, dietitian , lay  health worker,  nurse-led sessions,  psychologist-led  group sessions,  patient-led education,  and self-study  course	Physiological  Weight, lipids, blood  pressure, glycemic control  Behavioral  Dietary, physical activity,  self-care skills  Knowledge  Diabetes  Other  Economic measures; health  care utilization  Psychosocial  Attitudes, problem solving,  anxiety levels, quality of  life  QOL: yes, in 5 studies  Kaplan 1987-increase in  QOL at 18 months for  lifestyle intervention  (intensive counseling on  diet + physical activity).  Gilden 1992-increase in  QOL at 2 years follow-up  for coping skills  intervention (6 weekly  sessions +18 monthly  support group sessions).	Authors qualitatively  summarized outcomes to  generate hypotheses,  categorize variables for  quantitative syntheses, and  illustrate the vast  heterogeneity of methods  and outcomes in the  included literature.
	Notably, the authors  emphasize limited  literature measuring QOL  and long-term clinical  outcomes. Authors  advocate for a more  holistic view of patients,  including that QOL  outcomes take precedence  in future research and  more diverse study  populations.
		Content	Three studies (deWeerdt,  Glasgow, Trento)  considered “brief  interventions” with no  significant difference  between intervention and  control groups.
		Physical activity, diet,  nutrition, SMBG, weight  loss, barriers, social  support, foot care,  general DM knowledge,  self-adjustment of insulin,  goal setting, problem  solving, stress  management, patient  empowerment, self-  control, self-management
		Setting  Not specifically identified in  the review; described as  heterogeneous
Norris SL, Lau J, Smith SJ, Schmid CH, Engelgau MM. Self-management education for adults with type 2 diabetes: a meta-analysis of the effect on glycemic control. Diabetes Care. 2002;25:1159-1171.
1980-1999: MEDLINE, ERIC,  CINAHL; manual search  of relevant journals, and  experts were consulted  for citations	Included English-language  RCTs, DSME  interventions, and other  interventions  when delivered in  combination with  DSME (if effect of DSME  could be examined  separately).	Components  Didactic or collaborative  DSME, individual and  group education, support  groups, home visits,  dietician, flashcards by  lay health workers, meal  demonstration, feedback,  telephone follow-up,  psychologist-led group  sessions, weight loss  program, empowerment  techniques, computer-  assisted knowledge  assessment and  instruction	Physiological  GHb, HbA1, HbA1c  Psychosocial  QOL: not identified as a key  outcome of this review	Meta-regression indicates no  significant interactions  except total contact. In 15  studies, GHb  measurements were  reduced for every hour of  additional contact, which  approximates 23.6 hours  of contact between an  educator and patient to  achieve a 1% reduction;  brief intervention appears  to be less effective.
	Excluded abstracts and  dissertations.  N = 31 studies
		Content  Diet, physical activity, SMBG  or urine glucose, foot  care, coping, self-efficacy,  identifying and preventing  complications, goal  setting, and modeling  Setting  Clinic, home, and senior  center		Authors suggest that  psychosocial mediators,  cultural relevancy, and  health care system  structure and primary care  linkage may account for  the heterogeneity in  outcomes.
				Authors stress that further  research to identify  predicators and correlates  needs to focus on  psychosocial attributes,  social support, and  problem-solving skills.
Norris SL, Nichols PJ, Caspersen CJ, et al. The effectiveness of disease and case management for people with diabetes: a systematic review. Am J Prev Med. 2002; 22(suppl 4):15-38.
1966-12/2000: MEDLINE,  ERIC, CINAHL,  HealthSTAR, Chronic  Disease Prevention  database (health  promotion and education  subfile), Combined Health  Information Database,  diabetes, health  promotion and education  subfile), journals hand-  searched, reference list of  included articles and  consultation with team  experts for relevant  citations	Included published  comparative study  designs, English  language, conducted in  established market  economies as defined by  the World Bank, studies  with primary investigation  of disease (as defined by  review team) and case  management  intervention, reported  information on 1 or more  outcomes of interest  preselected by review  team.	Components  Disease or case  management along with  DSME, telemedicine  support, insulin  adjustment algorithms,  group support, visit  reminders, hospital  discharge assessment  and follow-up  Content  Not applicable or not  specified  Setting  Managed care organizations  and community clinics	Physiological  A1C, weight, BMI, BP, and  lipids  Behavioral  SMBG, patient health care  utilization; provider  screening, monitoring and  treatment): A1C, lipids,  dilated eye exams, foot  exams, proteinuria  Other  Health care system: health  insurance, provision of  services, health care  utilization (admissions,  outpatient visits, length of  stay), public health  services, economic  outcomes  Psychosocial  Self-efficacy, patient  satisfaction, quality of life	Authors note research gaps  in the areas of intervention  effectiveness on long-term  health and QOL outcomes,  as well as diverse  populations and settings.
				QOL Instrument: Peters et al  (1998) used the SF-36  with 2 (unspecified)  diabetes- specific  questions added.
	Excluded: studies  characterized as limited  quality based on the  number of threats to  validity, dissertations, and  abstracts.  N = 42 studies		QOL: Assessed in 2 of 42  studies (1 in adults and 1  in children); adult cohort  study reported relative  change of +4.7%  between groups.
Renders CM, Valk GD, Griffin SJ, Wagner EH, Eijk Van JT, Assendelft WJ. Interventions to improve the management of diabetes in primary care, outpatient, and community settings: a systematic review. Diabetes Care. 2001;24:1821-1833.
1966-2000: MEDLINE,  EMBASE, CINAHL, EPOC,  and Cochrane clinical trial  registries (1999);  reference list of selected  articles was reviewed	Included studies of  effectiveness of  interventions to improve  process of care or patient  outcomes among type 1  or type 2 that had (1)  randomized or quasi-  experimental trials, (2)  interrupted time series  with defined interventions  and a minimum of 3  before-and-after time  points, (3) nonrandomized  studies with a second  controlled site, and (4)  predetermined measures  of patient outcomes or  the process of health  care.	Components  Professional (education,  reminders, audit, and  feedback); organizational  (role revision, changes in  medical record systems;  patient education,  learner-centered  counseling, telephone  follow-up for missed  appointments); financial  (fee for services and  grants) or multistrategy  Content  Provider-guidelines  Patient-problem solving and  decision making  Setting  Primary care, outpatient,  and community	Physiological  Glycemic control, BP,  cholesterol, BMI, weight,  microvascular and/or  macrovascular  complications, albumin,  creatinine  Behavioral  Patient attendance, provider  process measures:  glycemic control, BP,  weight, microvascular  complications,  cholesterol, visits,  education, health survey,  compliance care  providers, albumin, urine  protein, creatinine,  hospitalizations	The authors note that  complex professional  interventions improved the  process of care, but  patient outcomes were  rarely assessed, making it  less clear to evaluate the  impact. Furthermore, the  authors emphasize that  measuring both process  and patient outcomes lead  to better understanding of  how to improve quality of  care.
	Excluded study interventions  classified as only patient  oriented.  N = 41 studies		Other  Hospitalizations, health care  system process  measures: glycemic  control, BP, weight,  cholesterol, visits,  education, microvascular  complications  Psychosocial  Well-being, quality of life  QOL: yes, but limited 3 of 41  (2 QOL and 1 well-being)  studies. QOL outcomes  showed no effect in 2  professional and  organizational  interventions; well-being  reported as positive effect  in 1 education  intervention.
Sarkisian CA, Brown AF, Norris KC, Wintz RL, Mangione CM. A systematic review of diabetes self-care interventions for older, African American, or Latino adults. Diabetes Educ. 2003;29:467-479.
01/1995-12/2000:  MEDLINE, HealthSTAR,  EMBASE, PsyclNFO,  Ageline, and Sociological  abstracts	Included English-language  published intervention  studies that described  intervention aimed at  changing knowledge,  beliefs or behavior;  targeted 1 or more of the  following: age >55 years,  African American or Latino  adults with DM; and  measure at least 1 of the  preidentified outcomes:  glycemic control, diabete-  srelated symptoms, or  quality of life.  N = 12 studies	Components  Group education plus  exercise class; group  sessions led by physician,  nurse, or dietitian;  didactic DM education  alone or in combination  with support groups,  individual counseling,  1-to-1 DM education,  bicultural community  health worker  liaison, nutritionist,  weekly pharmacist  appointments, grocery  store tour, follow-up  phone calls, physician  education	Physiological  Glycemic control and  diabetes-related  symptoms (established as  an outcome of interest by  review authors but not  assessed in any of the 12  included studies)  Psychosocial  Quality of life  HRQOL/QOL  In 5 of 12 studies: 4 of 8  RCTs and 1 of 4 pre/post  designs; 1 RCT reported  statistically significant  improved QOL in the  intervention arm (78  points vs 71 points);	Authors were unable to  summarize the effect of  self-care interventions on  quality of life because of  the heterogeneity of QOL  instruments.
				The authors suggest that  age- and culture-specific  interventions be designed  for clinical trials.
		Content  Exercise, nutrition, goal  setting, “standard DM  education curriculum,”  others unspecified/unclear  Setting  Noted as urban or rural	1 pre/post design  reported improved mean  scores on the QOL  instrument, with no mean  change in A1C at end of  intervention.
Steed L, Cooke D, Newman S. A systematic review of psychosocial outcomes following education, self-management and psychological interventions in diabetes mellitus. Patient Educ Couns. 2003;51:5-15.
1980-2001: EMBASE,  MEDLINE, Psyclit; hand  search of reference list of  reviews and retrieved  papers	Included pre/post or  controlled trial design  studies published in  English-language peer-  reviewed journals, adults  with type 1 or 2 DM,  provision of DMSE or  psychological intervention  and assessed either QOL  or psychological well-  being.	Components  Classified as education, self-  management, or  psychological biofeedback.  Content  General education, diet,  meal planning, goal  setting, reinforcement,  modeling, reward  systems, problem solving,  exercise sessions,  behavior change  maintenance, relaxation,  communication skills, and  social support  Professional or layperson-  led discussion groups  Setting  Not reported	Psychosocial  Overall psychological well-  being, depression,  anxiety, or emotional  adjustment and QOL  HRQOUQOL  Ten studies used a generic  measure, 9 used a  diabetes-specific  measure, 1 used both,  and 5 studies used an  overall psychological  well-being measure.	This review did not speak to  issues related to minority  or disadvantaged groups.  The authors were unable  to examine the efficacy of  different intervention  components; noted as the  original aim of the review,  multiple components in  single studies and the  overlap between  components in different  intervention categories  hindered these attempts.  Self-management  interventions had  beneficial impact on QOL.  None of the included  studies using psychological
	Excluded interventions that  provided only diet or  exercise or intensive  insulin regimens.  N = 36 studies		Generic QOL measures: 2 of  9 studies showed  improvement relative to  control; DM-specific QOL  measures: 6 of 9 found  improvement.  Overall psychological well-  being: 1 of 5 reported  significant improvemen.  Global and disease-specific  QOL were not assessed in  any of the psychological  interventions.
				Authors suggest that future  research should use  diabetes-specific studies  measures within larger  controlled trials.
				QOL Instrument: SF-36 and  DQOL were the most  frequently used generic  and diabetes-specific  instruments, respectively.
van Dam HA, van der Horst F, van den Borne B, Ryckman R, Crebolder H. Provider-patient interaction in diabetes care: Effects on patient self-care and outcomes: a systematic review. Patient Educ Couns. 2003;51:17-28.
1980-10/2001: MEDLINE  Advanced, EMBASE,  PsycIit/PsycINFO, and The  Cochrane Library	Included English-language  RCTs and quasi-  experimental studies  testing the effects of  modification of patient-  provider interaction and  provider consulting style  on DM self-care and  outcomes.	Component  Provider training,  questionnaire prompts,  feedback reports; patient-  automated telephone  calls, nurse telephone  feedback, individual  education, empowerment  group ed. sessions, group  consultations with doctor,  and individual  consultation with  standard DM education	Physiological  BP, lipids, BMI, GlyHb,  general health,  mortality  Behavioral  Health behavior  Knowledge  DM knowledge  Psychosocial  Satisfaction, well-being,  mental health, self-  efficacy, DM-specific QOL  scores	Methodological quality was  assessed using a modified  version of the Van Tulder  criteria list; the authors  include a table of the  criteria. Nineteen items  (questions) are categorized  into the following  categories: patient  selection, interventions,  outcome measurements,  and statistics. The mean  quality score was 17.3 out  of 19 total points.
	Excluded non-type 2 DM  patients, case, case  control, and  nonexperimental studies.  N = 8 studies
		Content  Provider-behavioral change,  communications, patient-  centered consulting,  guidelines; patient-diet,  self-care, problem  solving/barriers, decision  making  Setting  General practice or hospital  outpatient	Patient: communication,  interaction, lifestyle,  perception of care,  functional, attitudes  Provider: beliefs, attitudes,  and behavior change;  adherence to guidelines,  time spent with patient	Authors emphasize a shift  from the traditional  medical model and toward  a patient empowerment  model with shared  decision making.
			HRQOUQOL  In 2 of 8 studies:  1 with no change in HRQOL  scores and 1 with  “better” QOL.	Authors note they found no  evidence that the 3 most  promising studies have  ever been replicated  among diverse patient  populations or different  care settings, thus causing  alarm.

RCT, randomized controlled trial; DSME, diabetes self-management education; DM, diabetes mellitus; BP, blood pressure; FBG, fasting blood glucose; BMI, body mass index; QOL, quality of life; HRQOL, health-related quality of life; DQOL, diabetes quality of life; DCCT, Diabetes Control and Complications Trial Research Group; SES, socioeconomic status; PCP, primary care provider; CT, controlled trial; BGSM, blood glucose self-monitoring; DNS/NCM, diabetes nurse specialist/nurse case manager; SMBG, self-monitoring of blood glucose.