The impact of interventions on appointment and clinical outcomes for individuals with diabetes: a systematic review

Background Successful diabetes disease management involves routine medical care with individualized patient goals, self-management education and on-going support to reduce complications. Without interventions that facilitate patient scheduling, improve attendance to provider appointments and provide patient information to provider and care team, preventive services cannot begin. This review examines interventions based upon three focus areas: 1) scheduling the patient with their provider; 2) getting the patient to their appointment, and; 3) having patient information integral to their diabetes care available to the provider. This study identifies interventions that improve appointment management and preparation as well as patient clinical and behavioral outcomes. Methods A systematic review of the literature was performed using MEDLINE, CINAHL and the Cochrane library. Only articles in English and peer-reviewed articles were chosen. A total of 77 articles were identified that matched the three focus areas of the literature review: 1) on the schedule, 2) to the visit, and 3) patient information. These focus areas were utilized to analyze the literature to determine intervention trends and identify those with improved diabetes clinical and behavioral outcomes. Results The articles included in this review were published between 1987 and 2013, with 46 of them published after 2006. Forty-two studies considered only Type 2 diabetes, 4 studies considered only Type 1 diabetes, 15 studies considered both Type 1 and Type 2 diabetes, and 16 studies did not mention the diabetes type. Thirty-five of the 77 studies in the review were randomized controlled studies. Interventions that facilitated scheduling patients involved phone reminders, letter reminders, scheduling when necessary while monitoring patients, and open access scheduling. Interventions used to improve attendance were letter reminders, phone reminders, short message service (SMS) reminders, and financial incentives. Interventions that enabled routine exchange of patient information included web-based programs, phone calls, SMS, mail reminders, decision support systems linked to evidence-based treatment guidelines, registries integrated with electronic medical records, and patient health records. Conclusions The literature review showed that simple phone and letter reminders for scheduling or prompting of the date and time of an appointment to more complex web-based multidisciplinary programs with patient self-management can have a positive impact on clinical and behavioral outcomes for diabetes patients. Multifaceted interventions aimed at appointment management and preparation during various phases of the medical outpatient care process improves diabetes disease management.


Background
Diabetes is a complex chronic illness with significant health and financial implications. It has risen to epidemic proportions in the United States affecting approximately 26 million individuals in 2010 [1]. Projections reveal that if the current increase in diabetes incidence persists and diabetes mortality remains relatively low, prevalence will increase from the current level of 8.3 to 33 % of the adult population by 2050 [2]. Estimates indicate that the United States spent $218 billion in costs for pre-diabetes and diabetes care in 2007 [3]. The American Diabetes Association (ADA) and Healthy People 2020 propose guidelines and objectives for effective diabetes care management to reduce the incidence and economic burden of diabetes [4,5]. These objectives purport routine medical care with goals and treatment plans individualized for each patient, self-management education and on-going support to reduce the risk of diabetic complications [4].
According to ADA guidelines, which may vary from year to year based on evidence, people with diabetes should receive diabetes self-management education (DSME) at the time their diabetes is diagnosed and as needed thereafter. HbA1c test should be performed at least 2 times a year. The fasting lipid profile (total cholesterol, LDL, HDL, triglycerides) should be measured at least annually. A routine urinalysis and microalbuminuria test should be performed annually to assess nephropathy. A comprehensive foot exam should be performed every year to identify risk factors for ulcers and amputations. A dilated eye exam is recommended every year. Flu vaccines should be provided annually to all patients with diabetes. Pneumococcal vaccines are recommended for all patients over 2 years old. Self-monitoring of blood glucose (SMBG) should be performed three or more times a day for patients using multiple insulin injections or insulin pump therapy.
The percentage of United States adults with diabetes who received preventive care practices in 2009-2010 were as follows: ever attended diabetes self-management class, 57.4 %; check HbA1c ≥ 2 times a year, 68.5 %; annual foot exam, 67.5 %; annual eye exam, 62.8 %; annual flu vaccine, 50.1 %, and; daily self-monitor of blood glucose, 63.6 % [6]. Many factors including demographic, psychological, social, disease, treatment, provider, organizational, and care delivery related factors contribute to poor adherence [7]. These low levels of preventive care suggest an opportunity to enhance adherence to guidelines for effective disease management through appointment management and preparation because before diabetes preventive care practices can be instituted, patients must first be scheduled for and attend their provider appointments. Therefore, this study focuses on organizational and care delivery system related factors that relate to appointment management, as well as regular monitoring of relevant patient information integral to disease management.
Routine medical care starts with scheduling the patient with the provider for preventive care services. The patient can be scheduled for the next visit immediately after a provider visit or at a later time when the patient requests an appointment by phone or electronically. Interventions that proactively schedule the patient with their provider are a necessity for timely treatment decisions. Once patients are scheduled for their provider appointments the next step is to ensure that they attend their appointments. Studies show that no-show rates for diabetic patients vary from 4 to 40 % [8]. Literature also indicates that diabetic patients with higher no-show rates have poorer outcomes e.g., higher glycosylated hemoglobin (HbA1c) levels and poorer glycemic control than patients who attend appointments [8]. Without interventions to encourage patients to schedule and attend their provider appointments, other multifactorial interventions to reduce diabetes complications and costs of care cannot be initiated.
Research indicates that diabetes patients actively involved in their self-management experience improved Quality of Life (QOL) and improved HbA1c levels [9,10]. Currently, most diabetes care is provided in primary care practices. Accomplishing diabetes care objectives during fifteen to twenty minute appointments can be challenging for primary care providers. A provider cannot prepare individualized patient care without important patient information regarding self-monitoring blood glucoses (SMBG), daily diet and nutrition, exercise or physical activity, and medication information and compliance. To aid in the process of effective disease management, patients must take an informed and active role in the process. Interventions that aid the patient in communicating this information to the provider would expedite patient care delivery and allow the provider more time for individualization of the patient's treatment plan and patient support in self-management.
Literature examining interventions in diabetes care is extensive and offers a wide variability in types of interventions ranging from medication to web-based self-management tools with varying impact on diabetes outcomes. Different from the earlier literature reviews, the purpose of this literature review is to evaluate interventions that apply to appointment management and preparation, and determine their impact on appointment, clinical and behavioral outcomes for diabetic patients. This review examines interventions based upon three focus areas: 1) scheduling the patient with their provider; 2) getting the patient to their appointment, and; 3) having patient information integral to their diabetes care available to the provider. The hypothesis of this study is that interventions, which improve appointment management and preparation, are significantly associated with favorable appointment, clinical and behavioral outcomes.

Data source
This literature review was completed in February 2014. MEDLINE, the PubMed interface, was the primary database utilized. The following combination of MeSH terms was used for the search: "Diabetes Mellitus"[Mesh] AND ("Intervention Studies" [Mesh] OR "Internet" [Mesh] OR "Reminder Systems" [Mesh] OR "Appointments and Schedules" [Mesh] OR "Patient-Centered Care" [Mesh] OR "Registries" [Mesh] OR "Guideline Adherence"[Mesh]) NOT ("Diabetes, Gestational" [Mesh] OR "Pharmacological Processes" [Mesh] OR "Pharmacological Phenomena" [Mesh] OR "Transplantation" [Mesh] OR "Cardiovascular Surgical Procedures" [Mesh] OR "Heart Diseases" [Mesh] OR "Incidence" [Mesh]). Additionally, the reference lists of included articles and literature reviews were also examined for additional relevant articles. We searched CINAHL and found no additional articles. The Cochrane database was also searched and did not reveal other systematic reviews on this topic.

Data extraction
The comprehensive literature search generated 4111 articles (See Fig. 1). Studies excluding gestational, pharmacological process, pharmacological phenomena, transplantation, cardiovascular procedures, heart diseases and incidence reduced potential relevant articles to 2810. Articles were limited to those involving adults (19+ per PubMed), written in English and containing an abstract, which further reduced the total to 1308. Two reviewers reviewed the abstracts independently. All possible articles that could not be excluded were recorded in a table. Each study was marked as "relevant", "not relevant", or "maybe" based on the provided information in the paper and the goals for this systematic review. Once the reviewers prepared the tables independently, the decisions were compared and discussed in a meeting. Disagreement regarding inclusion of the article was reconciled through discussion with all other authors. Finally, by excluding articles that were not related to evaluation of an implemented intervention, the sample was reduced to 211 articles. Full texts of the 211 articles were retrieved and outcomes were evaluated independently by two reviewers according to structural, process, and outcomes measures [11]. One hundred and thirty four articles were excluded because they did not relate to the three focus areas: 1) on the schedule, 2) to the visit, and 3) patient information; the remaining 77 articles were included in this literature review. Disagreements regarding interpretation of data extracted from articles were reconciled through discussion with the authors. However, description of the types of interventions and outcomes were summarized and trended.

Results
The articles included in this review were published between 1987 and 2013, with 46 of them published after 2006. The following is a list of countries and the number of studies from that country included in the review: United States (43); South Korea (15); Netherlands (4); United Kingdom (3); Canada (3); Australia (2); France (1); Finland (1); Iran (1); Italy (1); Norway (1); Taiwan (1) and; Turkey (1). Thirty-five of the 77 studies in the review were randomized controlled studies. Table 1 is a summary of study designs and interventions used in each article included in this literature review. Appendix 1 provides detailed information about the interventions that focus on three areas of diabetes outpatient care delivery system: 1) scheduling the patient with their provider; 2) getting the patient to their appointment, and; 3) having patient information integral to their diabetes care available to the provider.
The reviewed articles evaluated the impact of interventions on several outcome measures. We divided the outcome measures into two types: clinical outcomes and behavioral outcomes. Clinical outcomes include the value of laboratory test results such as HbA1c, LDL, HDL, systolic blood pressure (SBP), diastolic blood pressure (DBP), total cholesterol, triglycerides, fasting plasma glucose, creatinine, 2-hour post meal glucose, and the value of clinical measures such as weight and body mass index (BMI). Given the importance of HbA1c in diabetes care, Table 2 includes only HbA1c results. All other clinical outcomes are provided in Appendix 2. In Table 2 and Appendix 2, we present the difference between the clinical outcome value at baseline and after the intervention (e.g., HbA1c at baseline -HbA1c at m months after the intervention) for both intervention and control groups. Where available, the p-values are presented for the difference between groups and the difference within the groups.
The behavioral outcomes, summarized in Appendix 3, include measures related to self-management (SMBG testing, physical activity, foot care, diet, nutrition, self-efficacy, quality of life, and patient satisfaction), attendance to outpatient visits for laboratory tests, vaccinations, primary care and specialty care, adherence to ADA guidelines (annual foot exam, annual eye exam, and processes of care), and acute care utilization (emergency visits, and hospital admissions). Since different measures or tools are used in different studies, we did not provide the numerical values for the changes in outcomes. For example, patient satisfaction is measured using different survey tools. The attendance to laboratory visits are measured using the number of laboratory tests within the next 6 months or 12 months after the intervention, or the percentage of patients who had the recommended laboratory tests within a year. For adherence to recommended laboratory tests, we included the tests considered in that study, and for vaccinations we presented the vaccinations.
In Tables 3, 4, and 5, we summarize the primary outcomes from the studies in Table 2, Appendix 2, and Appendix 3, based on if the interventions were directed at getting patients on the schedule, to the visit, or with the necessary patient information, respectively. Reference numbers of studies with significant outcome findings are bolded. In the following sections, we describe the most notable findings from these studies.

On the schedule
For the purpose of this literature review, an intervention that enables a patient to schedule a provider appointment or laboratory test meets criteria for 'on the schedule'. Review of the literature found limited research studying scheduling interventions as compared to diabetes intervention research pertaining to communication of patient information to the provider. The scheduling interventions, summarized in Table 3, included sending reminders to schedule a provider appointment or laboratory test, scheduling when necessary while monitoring patient information, and open access scheduling to provide same-day access. Although phone reminders were found to be effective for the most part to increase patient attended appointments, impact on clinic outcomes, as with other interventions in this focus group, were mixed and only a few studies discussed proactive appointment scheduling or management.
Grassroots interventions such as letter and phone reminders have been used to remind diabetic patients to schedule a provider appointment or a laboratory test. While the letter reminder, which asked the patients to call and make an appointment, improve the clinical outcomes including HbA1c, and SBP significantly in one study [12], it was not very effective in improving the clinical outcomes in other studies [13]. In a RCT, a letter from the provider was mailed to patients prior to their birthday with a self-care handbook, preventive care checklist and recommendations for routine monitoring and screening resulting in a significantly increased percentage of patients with an HbA1c test, percentage of patients with one diabetes-related provider visit, and percentage of patients with an eye exam within 6   RCT, parallel-group; Measurements: baseline, 4 and 9 mos.

Web-based; SMS
Exclusion: avoid tight glycemic control, another severe disease, cannot do SMBG, or other family member in trial. 22 Fischer et al. 2011 [13] Type 1 or 2 (based on ICD-9 codes) Mailed report cards: n I = 2728, n C = 2729; Printable report cards: n I = 2357, n C = 3100; Provider report cards: n I = 2893, n C = 2564; >17 yrs, at least one visit to clinic within 18 mos; Denver, CO; US.
Web-based; Decision support; Evidence-based guidelines 53 Meulepas et al.

[30]
Type 2 n I = 353 patients, n I = 51 providers; n C = 129 patients, n C = 27 providers; documented diabetes for > 4 yrs at start of study; The Netherlands Controlled, non-randomized, before/after study with delayed intervention in control group; Measurements: 1 yr before intervention and 2 years after.
Phone reminder 54 Meulepas et al. 2008 [31] Type 2 n I = 431 patients, n I = 23 providers; n C = 469 patients, n C = 28 providers; in the south of The Netherlands Controlled, non-randomized study, before/ after; Measurements: 1 yr before intervention and 2 years after.

Web-based
Exclusion: Did not speak English; had an emergency room discharge or hospital admission for a diabetes-related complication during the 6 months before recruitment; were homeless or did not have control of the given living situation; had significant cognitive impairment or psychological distress; had known substance or alcohol abuse.  RCT; Measurements: base line and 3 mos. Web-based; Phone reminder; Telephone data line; Automated performance feedback; SMS Exclusion: Severe diabetic complications; liver dysfunction with enzymes >2.5x normal, or renal dysfunction, diagnoses of heart failure, angina, myocardial infarction, or stroke, pregnancy or lactation. 77 Yoon and HS Kim 2008 [92] Type 2 n I = 25, n C = 26; ability to perform SBMG, access websites, and cellular phone with web access; university medical center, urban city of South Korea.

Web-based; SMS
Exclusion: severe illness, renal insufficiency with creatinine > 1.5 mg/dL or on insulin pump.
I intervention group, C control group     [14]. In another RCT, patients receiving a phone reminder to schedule an appointment 10 days following a letter reminder had significantly higher return rates for an annual follow-up eye exam than those patients who received only a reminder letter [15]. In a pretest/posttest study, phone calls made by medical assistants to schedule follow-up appointments with the primary care provider significantly improved glycemic control (reduced HbA1c levels) for the patients who returned for their follow-up visit [16]. In another study using RCT, phone calls to schedule an appointment with a pharmacist approximately one week prior to the physician appointment significantly improved compliance to ADA standards of care including percentage of patients who had A1c test, fasting lipid profile, foot exam and vaccinations [17]. An automated outreach call to non-adherent patients advising them to schedule an appointment significantly improved the percentage of patients with a provider visit and with HbA1c test for those patients who were successfully reached [18]. In a multi-center cross-sectional study, a phone call to reschedule after a no-showed appointment for a periodic provider visit resulted in significantly increased patient attendance to annual provider review, and those patients who attended their annual review had significantly lower fasting blood glucose [19]. Different than the studies that consider reminders to patients only, one study combined reminders to the patient with reminders to the provider [20]. In a RCT, faxed reminders were sent to the provider for patient overdue laboratory tests and letter reminders were sent to the patients with a warning of overdue laboratory tests. Even though the decrease in HbA1c and LDL of the intervention group when compared to control group was not significant, the number of emergency visits and number of hospital days per year were reduced significantly [20].
Comprehensive diabetes management programs that are used to monitor patient status can also be used to facilitate scheduling of patients for their provider visits. In a retrospective cohort study, the care coordinator regularly reviewed patient uploaded information such as SMBGs and *indicates significant findings with p-value ≤0.05; ↓=decrease, ↑increase NP p-value is not given scheduled provider appointments when appropriate, resulting in significantly decreased percentage of patients with at least one emergency visit and hospital admission [21]. In another RCT, a nurse reviewed self-management by phone at regular intervals, and a multidisciplinary care team provided both group visits every month for 6 months and individual visits after patient self-referral or referral by another care team member. The HbA1c levels and number of hospital admissions significantly reduced for the intervention group [22]. Open access, a scheduling strategy that offers same-day appointments for patients, can aide patients in scheduling a provider appointment and needed laboratory testing [23]. A drawback with this type of scheduling strategy is that the patient has the responsibility to initiate the next appointment at the appropriate time as specified in diabetes practice guidelines. If the patient forgets the timing of laboratory tests and provider visits, and the clinic does not send reminders to the patient for scheduling their appointments, open access scheduling might reduce compliance to diabetes management guidelines. One retrospective cohort study showed that open access scheduling was associated with significant decrease in HbA1c and urine microalbumin testing [23]. Even though HbA1c levels, and the number ED visits and hospitalizations did not change significantly with open access scheduling, the study suggested that scheduling process should be adjusted for patients with diabetes to improve diabetes processes of care (HbA1c, LDL, urine microalbumin testing) [23].

To the visit
Attendance to provider appointments and laboratory testing is a necessary component for implementation of diabetes preventive care. Interventions facilitating patient attendance to the scheduled provider appointments or laboratory testing meet criteria for the focus area 'to the visit'. Review of the literature found fewer studies discussing interventions to facilitate getting the patients to their provider visits as compared to diabetes intervention research pertaining to communication of patient information to the provider. The interventions that are used to improve attendance to the scheduled visits include letter, phone call, and SMS reminders, and financial incentives, as summarized in Table 4. Phone and mail reminders were the interventions most studied to facilitate patient appointment attendance with positive clinical outcomes. More studies are needed to determine if SMS and web-based appointment reminders and financial incentives can also improve provider visit attendance.
Our literature review showed that letter reminders to patients regarding lab appointment information were associated with significantly increased average number of HbA1c tests within the study period, number of patients who had HbA1c test within 6 months, and percentage of patients   [24][25][26]. In a RCT, letters recommending appropriate laboratory testing were automatically mailed quarterly to patients without HbA1c tests in the last six months or without LDL within the last twelve months resulting in significantly increased number of patients who had HbA1c test within 6 months and LDL test within 12 months [26]. Letter reminders one week before the scheduled provider appointment significantly increased the number of provider visits and reduced the number of hospitalizations in another RCT [27]. Phone reminders to patients regarding provider visits and laboratory testing resulted in improved HbA1c levels [16,[28][29][30][31]. One study showed that monthly phone reminders to patients in the intervention group regarding laboratory or provider scheduled appointments resulted in significantly decreased HbA1c levels and systolic blood pressure in the intervention group when compared to the control group [28]. Two studies where the medical assistant or the secretary called each patient before their scheduled appointment day to remind them of the appointment were associated with significantly decreased HbA1c [16,29] and LDL levels [29]. In two studies using a controlled, nonrandomized before/after design, a Diabetes Support Service (DSS) called patients in the intervention group to remind them of scheduled appointments for laboratory testing, foot exam, fundus photography and scheduled appointments with the dietician and diabetes nurse. The intervention was associated with a significantly increased percentage of patients with at least four HbA1c tests a year [30] and significantly lower HbA1c levels in the intervention group when compared to the control group [30,31].
Letter reminders combined with phone reminders of the date and time of the patient's provider appointment or laboratory test resulted in improved health outcomes [25,27,32]. One study showed that a recall card system and phone call reminding patients of their scheduled follow-up appointment resulted in significantly increased the percentage of patients who had HbA1c within the last 6 months and LDL tests within the last 12 months, significantly decreased the percentage of patients hospitalized in the last 12 months, and significantly increased the percentage of patients with foot exams and eye exams in the last 12 months [32].
Web-based programs associated with self-management can successfully remind patients regarding provider appointments or laboratory testing. A RCT used a web-based system to improve self-management education, and used emails combined with short message service (SMS) to send reminders one week before the follow-up visit, and to remind the time of the HbA1c test if it is more than three months overdue [33]. This web-based education management system combined with email and SMS reminders resulted in significantly decreased HbA1c and total cholesterol levels in the intervention group compared to control group [33].
Financial incentives used with other interventions have the potential to improve attendance to scheduled visits or needed lab tests. In a quasi-experimental study, a reminder letter was sent to patients for the completion of lab tests, and were offered and provided a gas card when the tests were completed [24]. The study showed that the reminder letter combined with a financial incentive increased the number of HbA1c tests significantly [24]. In another study, structured group visits facilitated by a diabetes educator were used as the main intervention [34]. A $10-incentive was provided to the patients for each group visit they attended [34]. Group visits combined with financial incentive achieved an overall attendance of 78.4 % to group visits, and significantly reduced SBP levels and number of ED visits per year [34].

Patient information
ADA, Healthy People 2020 and the Chronic Care Model recognize the primary importance and responsibility of the patient in self-managing their diabetes care and collaborating with their providers to set treatment and goals for improved health outcomes [4,5,35]. Interventions that aide the patient in communicating important information regarding SMBGs, daily diet and nutrition, exercise or physical activity, medication information and compliance, and patients' needs to their provider or health care team meet conditions for the focus area 'with patient information' (see summary of interventions and findings in Table 5). This focus area of the literature review provided the greatest number of research studies when compared to the other two focus areas, 'on the schedule' or 'to the visit'. Systems with routine monitoring of patient information, managing patient medications and supporting patients' goals whether web-based, SMS, or Electronic Health Record (EHR) with interfaced registry, consistently showed improved patient clinical outcomes. This literature review identified multiple studies using web-based diabetes management interventions with care manager feedback. In a RCT study, patients entered SMBG readings, exercise amounts, weight changes, blood pressure, and medication data via a web portal [36]. The study nurse monitored self-management changes, and contacted patients using email or chat to make recommendations [36]. The intervention resulted in significantly decreased HbA1c, systolic blood pressure and total cholesterol levels in the intervention group as compared to the control group who visited their provider for usual care [36]. In another RCT study, a nurse contacted patients biweekly for a 30 min video conference to review biometric data uploaded to the web-based self-management module and discuss patients' problems in managing the disease [37]. The intervention significantly decreased HbA1c levels in the intervention group [37]. Another study, which used randomized, singlecentered, controlled trial with parallel group design, evaluated a web-based program used by patients to review their online medical records, upload their SMBG levels, enter information about their exercise, diet and medication, and send secure emails to the care manager [38]. The care manager reviewed SMBG readings, guided health behavior, adjusted medications, and responded to patients' messages [38]. This web-based program, which provided ongoing tracking and documentation of patients' needs and care, decreased HbA1c levels significantly [38]. Seven studies combined web-based diabetes management program with SMS and were associated with significantly decreased HbA1c levels for the intervention group after implementation [39][40][41][42][43][44][45]. In six of those studies using quasi-experimental pretest/ posttest method conducted by the same research group, the nurse researcher reviewed uploaded patient data on the website, integrated patient clinical information into the patients' EHRs, provided education for self-management and sent weekly medication adjustment advice to the patient via SMS and internet [39][40][41][42][43][44].
Two studies showed that patients using a telephone data line to answer care coordinator's questions regarding daily SMBG readings, medication compliance and symptoms which were forwarded to the provider were associated with significantly increased quality of life (QOL) [46] and significantly decreased the percentage of patients with emergency visits and hospital admissions [21]. One study showed that patients receiving bi-weekly phone calls to review glucose and blood pressure readings had significantly reduced HbA1c and SBP levels [47]. Another study showed weekly phone coaching for goal setting and self-management significantly improved self-efficacy, diet, exercise, and foot care [48].
This literature review showed that the tools enabling decision support at the time of patient contact could improve compliance with preventive care services. A disease management application, which displayed trended electronic laboratory data linked to evidence-based treatment recommendations, resulted in significantly increased average number of HbA1c and LDL tests per year in a RCT study [49]. Patient data entered into Personal Digital Assistant (PDA), which enabled the tracking of evidence-based guidelines and provided reminders of due or overdue tests to providers at each patient visit, improved compliance to eye and foot exams [50]. The Diabetes Questionnaire and Reminder sheet, which is completed by the patient at check-in and reminded providers to check feet and update diabetes care flow chart used to document dates of preventive services in patient's chart, increased the number of HbAc1 tests, and compliance to eye and foot exams [51].
The utilization of an EHR driven diabetes registry within an integrated delivery system can improve diabetes health outcomes. A multicenter cross-sectional study showed that a computerized registration with templates for recording patient data from quarterly or annual diabetes visits integrated with patient's EHR resulted in significantly increased percentage of patients with HbA1c tests, and significantly decreased HbA1c, total cholesterol and triglycerides levels [19]. In a RCT study, a laboratory-based registry was used to fax and/or mail laboratory results, reminders of overdue laboratory tests, and quarterly population reports to providers, and to mail reminders for overdue tests and alerts for elevated test results to patients [20]. The integration of registry with patient and provider decision support decreased acute care utilization significantly, but did not decrease HbA1c level significantly [20]. A diabetes registry can be used to generate provider performance audits or provider patient panel reports to provide feedback regarding achievement of diabetes care guidelines including HbAc < 7.0 %. In three studies, these reports were shown to be associated with significantly improved diabetes processes of care (percentage of patients who had HbA1c test in the last six months, annual LDL cholesterol test, annual dilated eye exam, annual foot exam, and annual influenza vaccine) [26,52,53].
Personal patient held records summarizing goals, medical and laboratory outcomes for the year can assist both patients and providers as they organize individualized diabetes treatment plans. One study using clustered RCT showed that the intervention group utilizing patient-held health records resulted in significantly decreased HbA1c levels in the intervention group as compared to the control group [54]. However, web-based personal health records that allowed patients to review their medication lists, most recent test results and current treatments before the visit did not improve HbA1c levels in another RCT study [55].

Discussion
ADA and Healthy People 2020 recommended diabetic patients have routine laboratory tests and provider visits at regular intervals [4,5]. This literature review evaluated diabetes interventions, their effectiveness and resultant health outcomes, focusing upon the areas of scheduling the patient, getting the patient to their provider visit, and having patient information available to the provider. Figure 2 summarizes our findings by illustrating patient flow through the complex medical outpatient care delivery process with all potential interventions identified in this review. More specifically, Fig. 2 shows various components of diabetes outpatient care delivery, identifies phases of the process when interventions could be applied, identifies potential types of multifaceted interventions that could be utilized, and distinguishes whose responsibility it is for successful navigation through each phase of the care delivery system, e.g., provider and health care team versus patient.

Identifying gaps and highlighting future research opportunities
Diabetes management requires continuous monitoring and routine provider visits and laboratory tests [4]. This literature review showed that routine visits are either scheduled in advance or reminders are sent to patients to schedule their next appointment. When appointments are scheduled in advance, the attendance to scheduled visits might decrease as the lead time between the time the appointment is scheduled and the actual appointment time increases [56]. Therefore, advanced scheduling should be integrated with other interventions used to improve attendance to scheduled visits. In addition, clinics are moving from advanced scheduling to open access scheduling to reduce waiting times and improve access to care. However, one study showed that open access scheduling negatively affected the process outcomes for diabetes patients [23]. The mixed findings demonstrate the importance of provider or care team initiated interventions such as reminders sent to patients to schedule an appointment, or monitoring of patient information and scheduling when needed. The literature review showed that implementing automated or personalized phone reminders, which are relatively simple interventions and easy to employ by provider practices, are very useful in improving appointment making and attendance behavior. Web-based diabetes management tools are used to continuously monitor patient information and provide feedback to the patient. The continuously monitored patient information might include SMBG readings, patient medication use, blood pressure, weight, and nutrition or daily calorie intake. The degree of interaction with patients might range from providing feedback about SMBG readings by care manager to online coaching and structured counseling by diabetes specialist or nurse practitioner. The web-based systems can also be used to integrate laboratory testing and clinical information into patient's EHR. Web-based tools require patient, provider and care team involvement. Although a few of the studies discussed ease of use of the web-based interventions by patients and review patient satisfaction [46,57,58], none of the studies in this literature review discussed the ease of use for providers, provider satisfaction, or impact to the provider workload. Most of the web-based interventions using care manager monitoring and feedback used small sample sizes and did not discuss the direct and indirect costs and ease of implementation of the interventions for larger populations. More studies discussing provider workload and information regarding costs of the intervention may aid a practice in determining which inventions are most suited for their practice.
Selective financial incentives can improve quality of health services [59]. Three of the studies in this literature review incorporated financial incentives within the intervention. In one study, a gas card was given to patients after the completion of laboratory tests (HbA1c and LDL) and was associated with significantly increased laboratory testing when combined with a written reminder [24]. Another study discussed monetary incentives to providers for improving diabetes processes of care as demonstrated by significant increases in the percentage of patients with ideal glucose levels (HbA1c < 7.0 %) when combined with provider feedback and computerized reminders [53]. More research is necessary to determine the effect of both patient and provider financial incentives on patient health outcomes.
Tailoring the interventions according to patient population characteristics, needs, capabilities, and skills is an important factor that should be considered while choosing the set of interventions for implementation. For example, a web-based self-management program may not be as appropriate for elderly patients who may not be as comfortable with computer usage as a younger patient. The patients who have cellphones may not answer phone calls, but respond to SMSs. With the increasing use of smartphones, the patients might have regular access to email. However, underserved populations may not even have a regular phone or minutes to answer phone calls/SMSs in their cellphones. Changing technology and patient preferences with regard to contact/ communication should be considered when determining the future interventions to improve usage and effectiveness. Research evaluating the usage of interventions tailored for different patient groups is needed. Web-based tools with continuous monitoring can be used to categorize patients according to risk groups. Structured counseling and proactive scheduling of provider appointments might be used for high-risk patients to reduce the acute care utilization.
This literature review identified several interventions that improve appointment management and preparation. While impact of interventions on several clinical and behavioral outcomes is evaluated in these studies, effectiveness of interventions is not evaluated from a systems perspective. In other words, the interventions in the literature we reviewed appeared to be examined in isolation when they may, in fact, have repercussions throughout a provider's practice and patient population. Other factors such as ease of use by patients and providers, applicability of the intervention for larger populations and across other chronic diseases, and the cost of implementation are important concerns that may influence providers' decisions about adopting interventions in their practices. Research is needed that includes a more systematic view of the interventions and their implications beyond patient outcomes.
The methodologies used in the reviewed papers vary widely (including RCTs, quasi-experimental, pretest-posttest, retrospective cohort, non-randomized controlled trial, nested randomized trial, etc.). Even though RCT is considered as the best method in terms of strength and validity of the results, the reviewed studies that use other methods usually consider an intervention that can easily be implemented in large patient populations. These interventions include phone, letter/mail and SMS reminders to schedule an appointment or remind a scheduled appointment, and diabetes registries, and decision support systems to improve compliance to diabetes management guidelines. Since these interventions use large sample sizes, the included studies prove the applicability and impact of these interventions.
For the studies that consider using a web-based system with care coordinator feedback, RCTs are used with smaller sample sizes. Even though RCTs show the positive impact of such kind of an intervention, the small sample size might be an indicator of the difficulty of implementation due to the cost of the intervention.

Limitations
While the search in this literature review was conducted using several key databases and references were crosschecked, there may be publications not incorporated in the review because of the MESH terms used and inclusion criteria utilized. Only studies published in English were included which may create a chance for potential bias. All studies included in the literature review were peerreviewed publications. Although some interventions may be dated due to inclusion of studies published as early as 1987, less sophisticated interventions may have the same or better payoff and achieve similar goals at less cost and complexity in implementation.
One limitation of this literature review is that a metaanalysis was not performed due to inconsistency of the reported outcomes [60]. The included studies report a wide range of outcomes. Especially, the behavioral outcomes in Appendix 3, are not consistent. The measures related to self-management use different survey tools to assess patient satisfaction, quality of life, self-efficacy, etc. Other measures such as lab tests completed, vaccinations, provider visits, hospitalizations, and ED visits, are reported as either percentages or numbers (i.e. "percentage of patients who had ED visits" vs. "number of ED visits"). For clinical outcomes, the studies might report time effect, group effect, or time × group effect, which is again not consistent from study to study. Some studies did not use a control group or did not provide enough information before or after the intervention. This inconsistent reporting of wide variety of outcomes, and limited number of studies representing each outcome made the meta-analysis impractical.

Conclusions
The literature review showed that interventions from the simplest phone and letter reminder for scheduling or prompting of the date and time of an appointment to more complex web-based multidisciplinary programs with patient self-management can have a positive impact on clinical and behavioral outcomes for diabetes patients. Multifaceted interventions aimed at appointment management and preparation during various phases of the medical outpatient care process may provide a fail-safe against diabetes patients falling through the cracks of a fissured health care delivery system and maximize patient-provider limited time while obtaining the best possible disease management. While the overall results from this review suggest that interventions associated with appointment management and preparation result in better patient outcomes, an overwhelming absence of financial information in the reviewed studies may inhibit implementation. Indeed, practices may see an increase in costs associated with dedicated care managers and information technology support. Patients, and their insurers, may see an overall decrease in the costs of care when proper disease management is practiced. Unfortunately, these cost offsets may not be within the same cost center, and therefore, the providers paying for the interventions may not realize the cost benefits of enhanced patient well being. Future research must address these cost concerns and new policies may be necessary to ensure that interventions are beneficial for patients and providers.
This literature review also revealed that the trend of diabetes care is moving toward frequent monitoring of patient data and fluid management of patient diabetes care. Complex web-based systems are being overseen by an intermediate care manager, which may be an advance practice nurse, physician assistant or diabetes educator for 1) monitoring of SMBG levels, laboratory tests, medication compliance, diet and nutrition, physical activity and, 2) directing changes in patient care based on patient information. This intermediate care manager also directs the flow of patient information to provider, specialist and other members of the multidisciplinary health care team. The questions are whether the future of diabetes care and this type of continual monitoring will concentrate provider visits more toward those patients whose diabetes are not well-controlled or have a higher severity and what impact this change will have on overall diabetes outcomes. It seems reasonable that with the predicted increases in diabetes incidence and the already overloaded provider schedules that new strategies are needed to ensure access to care for all diabetes patients. Such strategies must include technical innovation that moves beyond the clinic visit, including continuous monitoring and risk assessment using emerging sensor technologies and smart algorithms, (semi) automated selection, execution, and tracking of interventions, learning algorithms to customize patient care plans, and gamification strategies to motivate and engage patient behaviors. Further, comprehensive cost-benefit analysis must become more widely accepted and practiced. The short and long term costs of interventions (capital, operational, maintenance, cyberinfrastructure, etc.) must be balanced against expected benefits from all stakeholder perspectives including patient access, outcomes, and satisfaction, clinic performance and provider utilization, inpatient usage patterns, reimbursement policies, and overall sustainability of the healthcare system. These strategies must be part of the larger, on-going efforts to transform healthcare delivery from being an uncoordinated assortment of specialties and special interests, supported by fee for service, to an integrated and holistic system that provides value to patients through prevention, early diagnosis, avoidance of chronic complications, and excellent therapy.   8 Carter et al. 2011 [37] On the schedule: NA Intervention group: Access to health education module with culturally age-appropriate education through videos and web sites, and social networking module linking intervention participants.
To the visit: NA Patient information: Self-management module: Nurse contacted patients biweekly for 30 minutes by video conference and reviewed uploaded data while patient viewed self-management video. Nurse and patient discuss data and behavior-change strategies. Patients would discuss problems in managing disease, e.g., medication side effects, and nurse provided feedback. Nurse transmitted data to patient's EHR (electronic health record). Provider transmitted updated treatment plans, lab results and other orders via portal to nurse and patient. To the visit: NA Patient information: Patients answered daily questions (changes in feet, blood sugar and medication) with Health Buddy (phone tool). Care manager using browser-based tool could automatically risk stratify information, forward patient information to provider, make provider referrals and reinforce self-management. 11 Cho et al. 2006 [57] On the schedule: Intervention and control patients scheduled for outpatient visits every 3 mos.
Both groups received diabetes management, nutrition, exercise, and blood glucose self-monitoring education.
To the visit: NA Patient information: Intervention group uploaded glucose levels, medications, BP and weight to web. Clinical instructors reviewed information daily and sent recommendations every 2 weeks. Medication changes referred to researcher and self-management or lifestyle changes referred to nurse or dietitian.   To the visit: NA Patient information: Both groups given mobile phone; SMBG levels, food intake, insulin dose, and activity levels automatically transmitted to server and made available to patient by web. Intervention group received real time clinical advice and structured counseling from diabetes specialist nurse. 22 Fischer et al. 2011 [13] On the schedule: Mailed patient report card reminding patient to schedule appointment if ≥ 2 mos since last provider visit.

NA
To the visit: NA Patient information: Point-of-care patient report cards generated automatically at visit (and mailed quarterly) included patient performance compared to national targets. Medical assistants encouraged patients' self-management goals. Quarterly provider performance report card generated from the registry. 23 Fischer et al. 2012 [69] On the schedule: NA The PRM system sends text messages to patients automatically according to an established schedule, and To the visit: Patients received text message appointment reminders 7, 3, and 1 days before appointments. Patient information: Patients received blood sugar reading requests every Monday, Wednesday, and Friday. The fasting blood sugar values outside the range of 70 to 400 mg/dL were automatically flagged in PRM and routed to a work queue. A registered nurse reviewed all flagged messages, contacted patients by telephone for follow-up assessment, presented out-of-range values to a physician, and ensured that both telephone encounters and patient-reported blood sugar measurements were appropriately documented in the medical record. 24 Glasgow                     Difference between groups, A Measured only after the intervention, NP p-value is not given