Healthy Outcomes

1996

Health Care Delivery
Targeting Clinical Guideline Evaluations
Meeting the Needs of Children with Special Health Conditions
Improving Diabetes Care in Minnesota
Partners for Better Health
Clinical Predictors and the Cost of Domestic Violence
The Over-Serviced and Under-Served
The Cost of Fractures

Women's Health
Mammography Screening and Early Detection of Breast Cancer
Attitudes Toward Mammography Screening
Patient Management for Pre-Labor Rupture of Membranes
The Mother-Baby Chemical Health Program

Health Care for Seniors
Depression in Elderly Patients
How Medicare Enrollees Feel About Their HMO
The Impact of Discontinuing "Seniors Plus"

Translating Research into Practice
Risk Adjusting for Medicare Beneficiaries in HMOs
The Nicotine Dependence Class
Improving Prevention Services

1994 / 1995 / 1996 / 1998 / 1999 / 2000 / 2001 / 2002 / 2003 / 2004

Investigators at the Group Health Foundation (GHF) are collaborating with the Institute for Clinical Systems Integration (ICSI), HealthPartners, and the University of Minnesota on a series of guideline impact studies. These studies evaluate whether the use of clinical guidelines is associated with improved care and reduced costs. Below are some of the interesting findings resulting from these efforts to date:

1. Results from the ICSI cystitis guideline study indicate a maintenance in quality care. However, the impact of the guideline was more noticeable with nurse-coordinated care than with physician-directed care. This was not unexpected because the clinics that were included in the study emphasized coordination of cystitis care by nurses.

2. The ICSI upper respiratory infection (URI) guideline reduced the use of antibiotics at first contact for patients with simple URI. The guideline also recommended telephone (as opposed to office visit) consultation of generally healthy patients. However, patients who were managed over the phone came in to the clinics later in large numbers, though most had no change in diagnosis. Thus, there were no significant cost savings associated with this guideline. The apparent lesson to be learned is that patient expectations are key to the success of some guidelines. This guideline may need revision and/or could incorporate a significant educational component for patients and employers to ensure its success.

3. The ICSI diabetes guideline focuses improved glycemic control. A study of HealthPartners members has shown that clinic and hospital resource use can be lowered for diabetic members through good glucose control. This finding indicates that, while the diabetes guideline will take some resources to implement, failing to successfully implement this guideline will be very costly.

4. The ICSI hypertension guideline is currently under evaluation to assess its impact on the use of recommended hypertension drugs and control of hypertension in primary care clinics.

5. To evaluate the impact of the ICSI adult low back pain guideline, GHF is conducting a pre- and post-guideline survey of more than 400 adults. Questions about pain, functional status, educational materials, medications, and satisfaction with care are being asked. Costs associated with guideline use and physician adherence to the guideline are also being studied.

6. The ICSI failure to progress in labor guideline evaluation is in progress. The impact of intervention with amniotomy and Pitocin on time to delivery, method of delivery, and maternal and infant outcomes will be assessed.

7. An upcoming study will assess the attitudes of female patients in HealthPartners toward the new ICSI guideline recommendation regarding frequency of Pap smear screening for cervical cancer. Women from various age groups and with varied screening histories will be surveyed to obtain their views.

8. From a health improvement and disease prevention standpoint, the ICSI tobacco cessation guideline is one of the most important. A study is just beginning that will survey patients in two clinics to learn whether implementing the guideline leads to more effective support for quitting and higher "quit rates." In addition, we hope to learn how satisfied tobacco users are with clinic efforts to help them quit. The results of this study should be very helpful in encouraging and shaping the implementation of this guideline.

In 1997, GHF will use its accumulated experience in guideline evaluation to develop improved strategies for guideline implementation.

Research Team
GHF/HealthPartners - Patrick J. O'Connor, MD, MPH; Andrew Nelson, MPH, Leif Solberg, MD, Cheri J. Rolnick, PhD, MPH; Raymond Boyle, PhD; William Rush, PhD; Todd Gilmer, MA; Jody Jackson, BA; Elaine Quiter, MS; Jerry Amundson, BA; Brita Hedblom, BA

University of Minnesota - Jon Christianson, PhD

Funding
Institute for Clinical Systems Integration

return to top

Project ACCORD (Advancing Child-Centered Outcomes through Relationship Development) is a collaborative effort between Group Health Foundation and the Minnesota Department of Health. This project is designed to track the process of guideline development and to monitor the implementation of guidelines into clinical practice. The goal of the project is to develop a guideline implementation model for conditions prevalent among children with special health care needs. Project ACCORD is examining the care provided to pediatric asthma patients within HealthPartners.

As part of Project ACCORD's activities, we have retrospectively documented the process used by the Institute for Clinical Systems Integration in developing the pediatric asthma clinical guideline. We have also documented several outcome assessment activities, including a special joint research project, The Pediatric Wheezing Survey, conducted by Park Nicollet Medical Center and Group Health Foundation. In order to gather more hands-on information about the processes being used to implement this guideline, we have conducted interviews with nurses, health educators, and physicians in a variety of clinical settings. Through this process, we hope to gather more detailed information about the factors that may help or hinder the implementation of clinical guidelines.

A major thrust of Project ACCORD is the development of public and private partnerships to support the optimal management of pediatric asthma. Of particular interest will be the identification of "wrap-around" service components, such as education, day care, and social services. These areas will be targeted in the development of a model that incorporates and coordinates support from organizations outside of the traditional medical to manage the child's special needs.

The remaining two years of this project will be devoted to the collaborative development of a generic guideline implementation model(s). Children with a variety of special health conditions who receive their medical treatment through managed care delivery systems will be targeted. The result of our efforts will be a useful quality assurance tool for health care delivery systems that serve this population, improving the experiences of children and their families as they access a range of services to help in managing their conditions.

Research Team
GHF/HealthPartners - Cheri J. Rolnick, PhD, MPH; Sharon Kaufer Flores, MS; Lloyd Fish, MD; Patrick J. O'Connor, MD, MPH

MN Department of Health - Jan Jernell, BS; Ann M. O'Fallon, BSN,MA; Nancy Vandenburg, PHN

Funding
Department of Health and Human Services, Maternal and Child Health Bureau

return to top

Group Health Foundation investigators have joined with the Minnesota Department of Health (MDH), the Centers for Disease Control and Prevention (CDC), and the Institute for Clinical Systems Integration (ICSI) to launch a diabetes improvement effort that will eventually be used as a state and national model.

1. The core activity included a survey of about 1600 HealthPartners members with diabetes. These members reported on the ways diabetes affects their lives, their views of health care providers, their current use of diabetes and preventive care practices, and their readiness to change health behaviors. Results show that members in both contracted and HealthPartners owned clinics were receiving excellent levels of diabetes care, and that members with more severe diabetes are receiving more intensive care.

2. A survey of HealthPartners primary physicians showed very high levels of awareness of current diabetes guideline recommendations such as those included in the new ICSI diabetes treatment guideline. Providers in both contracted and HealthPartners owned clinics appear to be putting forth their best to deliver recommended care.

3. A continuous quality improvement (CQI) intervention was developed at a volunteer clinic and is now being tested at three more clinics. The intervention was associated with a drop in glycosylated hemoglobin (long-term blood sugar level as measured by HbA1C) of one percent, with no increase in costs. A comparison clinic had a drop only .1 percent in glycosylated hemoglobin. Both clinics had a 25 to 30 percent increase in outpatient diabetes-related costs from 1994 to 1995, but the clinic using CQI had better health outcomes. The other clinic had little to show for the increased use of resources. This suggests that a CQI approach may improve both the efficiency and the outcomes of diabetes care.

4. A full-scale community trial of the CQI intervention to improve diabetes care is now beginning, with at least ten clinics participating. We will soon find out how powerful a tool CQI is for improving diabetes care, using the ICSI diabetes guideline as a road map.

5. From related ICSI-sponsored research on diabetes, it appears that patients with high glycosylated hemoglobin levels (i.e., poor control) have inpatient and outpatient costs that are approximately double those of patients with good diabetes control. Thus, while improving diabetes care is an energy consuming and challenging task, doing nothing to improve care may be even more costly in terms of human suffering and resource use.

6. Our research is zeroing in on the optimal number of glycosylated hemoglobin tests and clinic visits for adults with diabetes. Findings, which are still preliminary, appear to support the recommendations made in the ICSI diabetes guideline.

7. Improving diabetes care requires monitoring the outcomes of diabetes patients. A CDC-funded project will evaluate the accuracy of HMO databases and patient self-reports of eye exams and glycosylated hemoglobin blood tests diabetic members.

8. Three staff model clinics participated in a diabetes screening project. Patients who were at high risk for diabetes and had not received a diabetes screening in the previous year were invited to participate. Overall, 43 percent of patients responded to the screening; among those screened, only four new cases of diabetes and 11 cases of impaired glucose tolerance were detected.

9. In-depth interviews were conducted with all diabetic patients from one staff model clinic. questions were asked about how their diabetes affects their work, family life, and physical abilities; how they cope; how they take care of their diabetes; and how they view health care providers. The interviews are being analyzed to uncover ways to improve care for all patients with diabetes.

10. Much of the groundbreaking work on diabetes being conducted at HealthPartners and Group Health foundation has the potential for statewide use, an effort in which both the MDH and the CDC have expressed keen interest. Group Health Foundation investigator Patrick J. O'Connor is currently chair of the Minnesota Department of Health's Diabetes Steering Committee. In this capacity, he is an advocate for better diabetes care within and beyond the HealthPartners organization.

Research Team
GHF/HealthPartners - Patrick J. O'Connor, MD, MPH; Leif Solberg, MD; Dace Trence, MD; Sue Freeman, MD; Linda Cherney, RD, MPH; Theresa Pearson; Elaine Quiter, MS; William Rush, PhD; Laura Ellis, RN, BA

MN Department of Health - Don Bishop, PhD; Jim Bluhm, MPH; Cindy Clark, MA; Jay Desai, BA; Janel Harris, PhD; Sue Lasch, RN, NP-C; Laurel Reger, MBA

Funding
Centers for Disease Control and Prevention; Minnesota Department of Health, Institute for Clinical Systems Integration

return to top

Partners for Better Health: Improving Population Health

Group Health Foundation (GHF) investigators are working with the Center for Health Promotion in evaluating the impact of the Partners for Better Health (PBH) program on HealthPartners' members' health. PBH, a nationally recognized comprehensive program emphasizing prevention and early intervention, has set health improvement goals in several areas. The areas of interest include heart disease, maternal and infant complications, dental cavities, childhood injuries, breast cancer, diabetes, childhood immunizations, and domestic violence. The evaluation efforts of GHF include the following:

1. A survey of HealthPartners members conducted during the summer of 1994 established that members are willing to engage in health improvement activities with health care providers and health plan administrators to improve health. Members invited HealthPartners to do even more to support their efforts to lead healthier lives and increase healthy behaviors such as exercising, eating less fat, eating more fruits and vegetables, and not smoking.

2. During the summer of 1995, a survey of over 6,000 members established baseline levels of health status, behavioral risk factors, and readiness-to-change behavioral risks. Members with the highest risk of poor health outcomes were found to be the most ready to change their behavioral risk factors, and those with heart disease were most ready to take action to prevent further problems.

3. During the summer of 1996, a second follow-up survey was conducted with these same 6,000 members to determine their progress toward changing behavioral risk factors. Analysis will assess factors that predict behavioral change and the strength of the relationship between readiness to change and actual behavior change Member acceptance and response to critical elements of the initiative will be studied. In addition, analyses of the relationship between behavioral and disease related factors and costs of care will be conducted.

4. Special projects targeting heart include a study of the relative effectiveness of niacin and lovastatin for management of patients with cholesterol problems, with findings indicating that both drugs can be effective in certain groups of patients. Levels of physical fitness and readiness to change among patients with chronic diseases vary a great deal, and patients with heart disease fare the best. These findings further our understanding of what motivates patients to change, lead to program development improvements, and indicate the need for special efforts with members who have diabetes.

5. Efforts are underway to improve infant and maternal health. Data are currently being collected about low birth-weight babies and substance use during pregnancy.

6. In 1995, the PBH survey included three questions related to domestic violence.
Results indicate that as many as eight percent of surveyed members over age 40 did not feel safe in their personal relationships, with five percent who stated outright that they felt unsafe and three percent who stated that they "prefer not to answer."

The prevalence of members who felt "unsafe" differed by age: about six percent of members below age 70 did not feel safe, compared to over 12 percent of those over the age of 70. The reasons for this increase are not yet clear; one of the possibilities is more "prefer not to answer" replies among the older group.

Interestingly, the proportion of those not feeling safe increased with number of chronic conditions, from six percent of those with no chronic conditions and eight percent of those with one condition and nine percent of those with two or more conditions. Approximately equal numbers of men and women reported feeling unsafe. Among those with less than a high school education, 18 percent did not feel safe; this fell to 11 percent of those who were high school graduates and to five percent of those with more than a high school education.

The safety of elderly HealthPartners members is an area of concern. Most elderly members reported feeling comfortable discussing these concerns with their physicians, but few have actually been asked by their physicians about safety in relationships. This area may be one for increased clinical focus.

Research Team
GHF/HealthPartners - Patrick J. O'Connor, MD, MPH; Linda Cherney, RD, MPH; William Rush, PhD; Cheri J. Rolnick, PhD, MPH; Catherine Wisner, PhD; Nico Pronk, PhD

Funding
HealthPartners

return to top

This study identified clinical indicators of domestic violence and the added cost of providing health care to victims of domestic violence. Female victims of domestic violence continuously enrolled in the HealthPartners during 1994 were identified by referrals for treatment related to domestic violence between January of 1992 and December of 1994. A random sample of 1007 women were identified from the same time period to achieve a sample approximately equivalent to the estimated prevalence of domestic violence in the HMO population. All data were abstracted from computerized database in HealthPartners.

Significant differences were found between victims and the general female HealthPartners population with victims being younger and having had more visits for injuries and depression. The difference in number of visits for anxiety approached statistical significance. Age and domestic violence were found to be significant predictors of cost, with the estimate of added cost for domestic violence to be $1775 per year (120 percent). Breakdowns of costs between the groups revealed significantly higher numbers of general ambulatory clinic visits, out-of-plan referrals, affiliate visits, and mental health visits for victims of domestic violence.

The results of this study fill a gap in the scientific data on domestic violence, and demonstrate the impact of this issue in adding to the cost of health care.

Research Team
Catherine Wisner, PhD; Tod Gilmer, MA; Brita Hedblom, BA, Therese Zink, MD; Peggy Trezona, MS, RN, CS

Funding
Group Health Foundation

return to top

How should an HMO care for patients with hidden mental health conditions that complicate their medical conditions? These patients, who often are subjected to costly and invasive treatments that do nothing for them, can be said to be over-serviced and under-served.

The goal of the Primary Care / Mental Health Collaboration Project is to design a new paradigm for integrating health and mental health care. This project, which entails a partnership between Bristol-Myers Squibb and HealthPartners, has a special focus on over-serviced / under-served patients. The first phase of the project was a baseline evaluation of the health psychology program and other mental health services for primary care patients at HealthPartners owned clinics.

About ten years ago, HealthPartners introduced the health psychology program as an innovative model to facilitate collaboration between mental health and primary care. PhD-trained psychologists provide consultations to medical providers and counseling to patients-for example, helping patients to manage chronic conditions, such as headaches, depression and diabetes. Currently, eight health psychologists are working in seven primary care staff model clinics, as well as in selected sub-specialty departments.

Focus groups were conducted with the following four groups:
· health psychologists
· front-line clinic staff (RNs, LPNs, and receptionists)
· physicians from clinics with health psychologists
· physicians from clinics without health psychologists

Insights from the focus groups include:

The Health Psychology Program - Some physicians have developed effective working relationships with the health psychologists. However, there is a lack of visibility-so many physicians don't think of a referral to a health psychologist. Physicians also say it takes too long to get an appointment and patients sometimes resent the implication that the problem is "just in your head." Moreover, doctors lack sufficient time to try something new and the culture is not set up to facilitate collaboration.

About Mental Health Services - There appears to be a lack of communication and coordination between primary care and mental health. There also is a perception of access problems-long wait times, or lack of availability all together. Moreover, cultural differences between medical providers and mental health providers make consultation more difficult than with other medical specialists.

The Future of Integrated Care - There is a new initiative at HealthPartners to expand the concept of integrated care, building on the experience with the health psychology program. A physician commented about the future of this initiative: "I think there needs to be a real commitment from somewhere upstairs for this to happen. But it's not going to happen in six months. I think it's going to take several years, probably, to build up a trust level and a working relationship between primary care and mental health." The availability of adequate resources will be a critical factor in the success of future efforts to expand integrated care models at HealthPartners.

Research Team
GHF/HealthPartners - Lucy Rose Fischer, PhD, Thomson Davies, PhD; Steven Lucas, MD; Richard Heinrich, MD; Macaran Baird, MD; Todd Gilmer, MA; Adrew Nelson, MPH; C.J. Peek, PhD; Sondra Otto, MA

Funding
Bristol-Myers Squibb; HealthPartners

return to top

How many older women and men have fractures? What is the impact of fractures on the cost of health care services? This study, with data from HealthPartners enrollees, looked at the incidence and cost of fractures in men and women age 50+. For those between 50 and 64 years of age, women were approximately twice as likely to experience fractures as men-about nine in 1,000 women, compared to five in 1,000 men each year. After age 65 the incidence rates are much higher-about two in 100 women and one in 100 men.

Table 1 compares the cost of health care for a one year period before and after the fracture. Women enrollees between the ages of 50 and 64, and over 65 who had fractures were compared with a sample of women enrollees without any fractures, matched by age. As Table 1 indicates, women who experienced fractures generated somewhat higher costs than the matched sample of women in the year before their fracture. This suggests that women who experience fractures may be somewhat less healthy than other enrollees. Even so, the fracture itself has a very large impact on health costs. As Table 1 illustrates, the cost of health services more than doubles for women and men in both age groups in the year following a fracture. In contrast, the cost for the matched sample of women is essentially the same in the "before" and "after" years.

Analyses found no significant differences in self-reported health status (independent activities of daily living) by type of fracture (central or appendage) when compared to no fracture. Further analyses found significant relationships between numbers of chronic conditions and self-reported health status.

Research Team
GHF-Catherine Wisner, PhD; Lucy Rose Fischer, PhD; Gerald Amundson, BA; Brita Hedblom, BA

Merck Pharmaceuticals, Inc.-Tom Abbot, PhD; Mark Berger, MD

Funding
Merck Pharmaceuticals, Inc

return to top

As part of the Partners for Better Health Initiative, a study was conducted
to determine whether a difference could be found between the mammography screening histories of women diagnosed with early-stage and those diagnosed with late-stage breast cancer. The study was designed to assess the potential for reducing late-stage diagnosis of breast cancer by improving rates of mammography screening.

All women who received a breast cancer diagnosis between January 1990 and December 1994 were identified from a computer database. Each cancer case was categorized as Stage 0, I, II, III or IV, based on tumor registry records. Every woman with a late-stage (Stage III or IV) breast cancer at the time of diagnosis (N= 50) was included in the case group. Each case was then matched to an early-stage (Stage 0 / I and Stage II) control, based on year of diagnosis and age. A physician reviewed medical charts to determine the dates of all mammograms performed in the previous five years, the reasons each mammogram was performed, breast cancer risk factors, and treatment choices.

Table 1 provides a breakdown of subject age (both the percentage of women and the rate per 100,000 women with late stage cancer), and the number of late-stage cases diagnosed per year. The highest percentage (35 percent) of late-stage breast cancer was found in women between the ages of 40 and 49. As expected, the rate per 100,000 women increased with age. However, the rate per 100,000 women among 40 to 49 year olds was similar to that of women aged 50 to 59. The number of late-stage cases diagnosed was similar, ranging from 8 to 12 women per year.

Information indicating whether initial suspicion of a potential breast problem was the result of concern from the woman, the provider, or the results of a mammography were tabulated (Table 2). Women with late-stage breast cancer were more likely to be suspicious that something was wrong before seeing their provider or having a mammogram; 87 percent reported that they first raised the suspicion of potential cancer, compared to 57 percent of women with Stage II cancer and 24 percent of women with a Stage 0 or I diagnosis. Those with early-stage disease were most likely to learn of cancer based on mammography results (61 percent).

Matched pair analyses were used to analyze the differences between women with Stage III / IV and those with Stage 0 / I breast cancer. Late-stage patients were significantly less likely than their Stage 0 / I counterparts to have had a screening mammogram within 13 months of the mammogram which led to the cancer diagnosis. There was no significant difference in screening between Stage III / IV patients and those diagnosed with Stage II breast cancer.

Screening mammography has the potential to improve early identification of breast cancer for many women. It was of interest to note that the five-year incidence rate of late-stage breast cancer in our population was the same for women in their 40s as for those in their 50s. Current recommendations regarding annual breast cancer screening were derived from studies conducted at a time when overall mammography use was low, and the benefit of mammography screening was easily demonstrated in the age groups at highest risk of cancer. In the current era, with its "epidemiologic transition" in breast cancer incidence rates, the question of who should receive regular mammography screening must be reconsidered. The data presented here are provocative and suggest that the justification for screening 40 to 49 year old women may be as strong as the justification for screening 50 to 59 year old women.

Further reduction in the incidence of late-stage breast cancer may also be obtained from more thorough screening of women aged 50 to 75 years. The odds of women with early-stage cancer having had a screening mammogram were six times higher in this case-control study than the odds of women with late-stage. The data also reiterate that older women (over age 75) have the highest rate of late-stage breast cancer of any age group, and suggest that care providers need to consider regular mammography screening for these women. The effort to decrease late-stage diagnoses is increasingly important as the cost of treatment of late-stage cancer escalates and the rates of late-stage cancer are publicly reported. Gaining a better understanding of who, when and how to screen can have a significant impact on reducing morbidity and mortality related to breast cancer.

Research Team
GHF/HealthPartners-Cheri J. Rolnick, PhD, MPH; Patrick J. O'Connor, MD, MPH; Jody Jackson, BA; Raymond Boyle, PhD; Nico J. Pronk, PhD, Linda Loes, MD

Funding
HealthPartners Center for Health Promotion

return to top

Why do some women choose to get a mammogram, while others never seek out breast cancer screening? As part of the Partners for Better Health program evaluation, fifty women age 50 and older were interviewed regarding their attitudes toward mammography. Women from two clinics were identified from data obtained in a general health survey, and were categorized into one of four groups:

· women who had a mammogram in the past year
· women who had a mammogram between one and five years ago
· women whose last mammogram was more than five years ago
· women who had never obtained a mammogram

All interviews were audiotaped and transcribed. Questions focused on the reasons women do or do not go for screening, whether the subject knew someone with breast cancer and if this affected her decision to seek mammography, and what HealthPartners might do to encourage screening. Comments from women who obtained routine, annual mammograms were then compared with answers from women who had never had a mammogram.

The most common reason stated for getting a mammogram was that the woman's provider had recommended it. The most common reason for not seeking breast cancer screening was that woman did not feel it was necessary; these women stated that they felt healthy or had no family history of the disease. Women who regularly obtain screening suggested that others who avoid mammograms might fear pain-interestingly, most of these women did not report that their most recent mammogram was painful. Those in the "never" group suggested that they thought women might fear the results of mammography. About half of the women interviewed said that knowing someone with breast cancer had an impact on their screening decision, most often (95 percent) serving to motivate them to obtain screening.

Suggestions regarding ways to make it easier for women to get mammograms included combining the mammogram with other appointments, making mammograms available at more convenient locations, using alternative technology that reduces squeezing and discomfort, extending clinic hours, and making reminder calls. Most of those who had never had a mammogram said that there was nothing that could be done to make it easier for them to get one. Clearly, there is a need to design targeted strategies so more women will avail themselves of this important screening.

Research Team
Cheri J. Rolnick, PhD, MPH; Patrick J. O'Connor, MD, MPH; Jody Jackson, BA; Shirlee Sherkow, PhD; William Rush, PhD; Raymond G. Boyle, PhD

Funding
HealthPartners Center for Health Promotion

return to top

Pre-labor rupture of membranes (PROM) at term in a woman with an unfavorable cervix (dilatation of less than five centimeters) occurs in about ten percent of pregnancies. When it occurs, PROM presents a difficult management problem for obstetricians. Various treatment options exist; the provider can choose to use observation, administration of Pitocin, application of intracervical gel, or they may attempt a combination of these strategies. All methods of intervention have been supported in the scientific literature.

To determine the practices most commonly being used by HealthPartners providers, a retrospective chart review was conducted. This review allowed us (1) to document the incidence of favorable and unfavorable cervix, the incidence of spontaneous labor in each group, and the various augmentation strategies used when spontaneous labor did not occur, and (2) to track maternal and neonatal outcomes associated with each management strategy.

A computer search was conducted to identify all women with spontaneous rupture of membrane (SROM) seen at Fairview Riverside Hospital between January 1, 1992 and August 31, 1993. To be eligible, patients had to have an uncomplicated single-infant pregnancy with a headfirst presentation and a gestational age between 37 and 42 weeks. Women with a previous Cesarean section were included. Using these eligibility criteria, 438 women were identified.

Demographic characteristics of patients in each intervention option were compared. Analyses of various methods of management were conducted to examine five major outcomes:

· time from SROM to delivery
· delivery method
· infections in the mother or in the infant
· one-minute Apgar scores
· transfers to special nursery care

Analyses were conducted on the entire group of women and then each variable of interest was examined based on favorableness of the cervix. Those women whose cervical status was not documented in the chart were dropped from the latter analyses, leaving a total sample of 385. There were 295 women (77 percent) classified with a favorable cervix and 90 considered to have an unfavorable cervix.

The mean maternal age for the group was 30 years; mean gestational age was 39.4 weeks. There was no difference in either of these variables based on management method. Over 79 percent of the women had a favorable cervix and in 71 percent of deliveries, Pitocin was used.

Overall, the management approach appeared to have no significant association with C-section rates, infant Apgar scores, or transfers to special nursery care. The management approach was, however, associated with time from SROM to delivery. The use of Pitocin increased the probability of delivery twofold for all women and for those with favorable cervix. This association was not found for women with unfavorable cervix. The use of gel was negatively associated with time to delivery-women who received gel had a longer time to delivery than those who did not. This finding is most likely due to the fact that women who were not progressing were administered gel, not that the gel postponed their delivery. The less favorable the cervix, the greater the time to delivery.

Having a previous vaginal birth decreased both the time to delivery and likelihood of a C section delivery. Number of cervical checks increased the probability of infection, low infant Apgar score, and transfer to special nursery care; women were 1.3 times as likely to develop an infection for each check of the cervix. Women were approximately twice as likely to have an infection if they had a prior birth.

For the entire group of women, infant transfer to special nursery care was associated with the number of cervical checks and infection, and was inversely related to prior vaginal birth. When an infection was present in the mother or the infant, the infant was four times as likely to receive a low Apgar score and was nearly ten times as likely to be transferred to special nursery care. In addition, there was an association with parity; women who had previous children were twice as likely to have their infant transferred to special nursery care.

A provocative finding was that the use of gel appeared to be protective. None of the women with an unfavorable cervix who received gel had an infection, nor did their infants. Of the 90 women thus classified, ten were observed but received nothing to stimulate labor. Fifty-seven of the women were given Pitocin, 12 received gel and 11 were given a combination of the Pitocin and gel. Of the 23 women who received gel none became infected. Of the 57 women who received Pitocin alone, eight had infections.

By understanding patient outcomes, we can determine when intervention is beneficial and develop more effective management strategies to care for women during childbirth.

Research Team
GHF/HealthPartners-Cheri J. Rolnick, PhD, MPH; Barbara Hyer, MD; Todd Gilmer, MA; Jody Jackson, BA; Sharon Kaufer Flores, MS

Funding
Group Health Foundation

return to top

Although there is increasing recognition of the problems caused by substance use and abuse during pregnancy, there is a paucity of reports in the literature describing multidisciplinary, consultative services dedicated to early identification and intervention of those at high risk. The Mother-Baby Chemical Health Program was developed in 1988 to focus on pregnant women who use and abuse substances, and has had an important and positive impact.

The program established the following goals:

· Seventy percent of the patients will remain drug-free during pregnancy
· Ninety percent of the participants will deliver at 37 or more completed weeks of gestation
· Ninety percent of the infants will weigh >2500 grams (5.5 pounds) at birth

To achieve these goals, the program works to provide comprehensive health care and education to substance using/abusing pregnant women and their infants. The program provides interdisciplinary care in the areas of chemical health, obstetrics, and pediatrics. It provides these women and their infants a safe, non-judgmental environment that is designed to meet their physical, psychological, educational and social needs. At any point in time, the nurses were following 80 to 90 patients prenatally.

There were 352 women seen from January 1990 and December 1992; these women gave birth to 356 infants. Nearly 93 percent of the women delivered at 37 or more weeks gestation, and over 91 percent of the infants weighed >2500 grams at birth. Data from 1993 through1995 have not been officially tallied; preliminary information from 1994, however, indicates that 119 women entered the program. Unofficially, 97 percent reported no use of drugs during their pregnancy and approximately 89 percent of their infants weighed >2500 grams at birth.

The program has just begun its seventh year, with two nurses following approximately 45 patients each. If you are interested in obtaining a copy of a published article describing the program and more detailed results, contact Peter M. Mark, MD, HealthPartners Riverside Medical Center, Department of OB/Gyn, 2220 Riverside Avenue, Minneapolis, MN 55454.

Research Team
GHF/HealthPartners-Peter M. Mark, MD; Antoinette St. Pierre, RN, BS; Roxanne Michelson RN, MS, CPNP; Lawrence Condon, MD; Andrew Nelson, MPH; Cheri J. Rolnick, PhD, MPH

Funding
HealthPartners

return to top

As part of the Seniors Quality of Life Study (SEQOLS), 3410 men and women age 65 and older were screened using the Geriatric Depression Scale (GDS) to obtain an assessment of self-reported depressive symptoms. Provider recognition of the possibility of depression was assessed by the presence of one or more of the following clinical indications:

· a diagnoses of depression in any patient encounter
· one or more visits to a mental health specialist
· antidepressant drug treatment

Of the 3410 subjects, 16 percent (N=559) had an indication of depression based on the GDS; of these, half had an indication in their medical record of the possibility of depression. An examination was conducted to determine factors that might be associated with provider recognition of depression. Variables examined included gender, age and severity of depression based on the GDS. The variable most likely to result in provider recognition was severity. Providers were also more likely to recognize depression in women, regardless of age or severity of GDS score. Men, while less likely to be generally identified as depressed, were even less likely to be identified if they were between the ages of 65 and 74, or over age 85.

Although the GDS is a useful screening tool for determining the possibility of depression, it is not the same as a physician diagnosis. It may be that not all of those categorized as having self-reported depressive symptoms were truly in need of clinical recognition. However, the results of this analysis suggest that some depression is being overlooked, and confirm findings in the literature that clinical depression is often undiagnosed and untreated. By improving the ability of providers to identify elderly patients with depression, it may be possible to improve the quality of life and reduce health care costs for these patients.

Research Team
GHF/HealthPartners- Cheri J. Rolnick, PhD, MPH; Lucy Rose Fischer, PhD; Richard Heinrich, MD; Jody Jackson, BA

University of Minnesota, School of Public Health, Institute for Health Services Research-Judith Garrard, PhD; Nicole Nitz, Lori Luepke

Mayo Foundation-Cynthia Liebson, PhD

University of Minnesota, Office of Information Technology-Patricia Bland, MA

University of Minnesota, School of Public Health, Division of Biostatistics-Lance Waller, PhD

Funding
Agency for Health Care Policy and Research

retun to top

In January of 1996, two focus groups were conducted with Medicare recipients enrolled in HealthPartners to collect information about their experiences and perspectives. The findings from these focus groups have been used to develop a questionnaire for a national survey of Medicare enrollees in HMOs.

Participants in the two focus groups were recruited randomly from two categories of HealthPartners Medicare enrollees-persons age 65+ continuously enrolled for at least one year and persons age 80+ enrolled for at least two months. There were eleven participants in each focus group.

Overall, the participants in both focus groups expressed high levels of satisfaction with clinics, their doctors, and the managed care system. Most participants reported receiving personalized care at their HMO. However, there was some concern that health care was changing and becoming more like a "business."

Many enrollees felt that "you have to be your own advocate with your health care providers and in the health care system," even in a high quality system. The participants identified a need for self-advocacy in many aspects of care-monitoring medications, receiving referrals to specialists, and ensuring that personal preferences are respected.

Some of the participants complained about the following (although others in the groups reported positive experiences in these areas): not receiving test results; not being informed when their doctor had left or was no longer seeing primary care patients; not receiving reminders about preventive care as much as in the past; not having sufficient review of multiple medications; and not receiving referrals to specialists. A few potentially serious problems in diagnosis and treatment were reported. Even enrollees who had negative experiences, however, described their care at the HMO as generally very good.

Research Team
Lucy Rose Fischer, PhD; Sondra Otto, MA

Funding
Subcontract with Mathematica Policy Research, Inc. for the Physician Payment Review Commission.

return to top

On December 31, 1994, Seniors Plus, the Group Health Social HMO, was discontinued due to large financial losses. A research study was undertaken at GHF to assess the impact on frail elderly and their caregivers after losing long-term home care.

In June of 1995, a mailed survey was administered to approximately 500 former Seniors Plus enrollees who had received the home care benefit in 1994. Telephone interviews were conducted with a sub-group of frail elderly who said they were currently receiving less home care.

Many of these at-risk elderly were paying more and getting less care. For example, one woman used to have a daily home health aide who helped her exercise. Her daughter took over much of the care but did not have the time to do all of the exercises with her mother. The daughter reported: "I can tell there's a difference. She seems to be stiffer…just not as able as she was before."

About half of the interview sample switched from agency care to p rivate pay care, which is less costly. The discontinuance of Seniors Plus led to an increase in burden and stress form many of the family caregivers. Care coordination was designed as a key component of the Social HMO. Many elderly and their caregivers missed the supervision of the care system and the personal attention.

Research Team
GHF/HealthPartners-Lucy Rose Fischer, PhD; Annice Miller, MBA; William Rush, PhD; Jeanne Ripley; Catherine Wisner, PhD; Tom VonSternberg, MD; Andrew Nelson, MPH

Brandeis University-Walter Leutz, PhD

Funding
Group Health Foundation

return to top

The current method used by the Medicare program to determine the rate that HMOs are paid for services is based upon calculations of cost within fee-for-service settings. The results are often seen as inequitable; Medicare beneficiaries in some areas receive generous additional benefits without additional premiums, while in other areas, beneficiaries receive limited benefits with substantial out-of-pocket costs.

In an attempt to address this apparent inequity, and to help the Medicare program take advantage of changes in the employer group market, HMO Research Network members from Group Health Foundation, the Center for Health Studies (CHS) at the Group Health cooperative of Puget Sound, and the Center for Health Research (CHR) at Kaiser Permanente in Portland, Oregon, have been collaborating for two years on the Global Risk Adjustment Models (GRAM) Development Project. This effort has resulted in the development of one of the first comprehensive risk assessment models designed to predict future annual health care costs for all HMO enrollees, from newborns to the elderly.

Although the model has multiple uses, it is intended to demonstrate the feasibility of paying risk-based Medicare HMOs based on the difference in expected cost of Medicare beneficiaries compared to non-Medicare beneficiaries, within HMO cost structures. Under the direction of CHR research Dr. Mark C. Hornbrook, cost and morbidity data have been collected on a random sample of health plan members from MedCenters and HealthPartners of Minnesota, Group Health Cooperative of Puget Sound, and Kaiser Permanente in Cleveland and the Northwest. The model will now be tested through pilot implementation at Kaiser Permanente in Portland, Cleveland, New York, and Denver.

Research Team
CHF-Michael Goodman, PhD; Andrew Nelson, MPH; William Rush, PhD; Catherine Wisner, PhD; Todd Gilmer, MA

Center for Health Studies-Paul Fishman, PhD

Center for Health Research-Mark Hornbrook, PhD; Merwyn Greenlick, PhD

Funding
Subcontract with the Kaiser Center for Health Research for the Health Care Financing Administration

return to top

Following the PROSTEP nicotine patch study, Dr. David Klevan has continued to lead educational sessions designed to acquaint participants with the use of the patch. The 90-minute classes are held at Como clinic and are taught by both Dr. Klevan and a nurse. The sessions focus on smoking cessation motivation and relapse prevention. Attendees view a video, receive educational pamphlets, and are given a free seven-day supply of the patch and a one-month prescription for nicotine gum or the patch.

Between January 1994 and December 1995, 215 people attended the educational sessions. In May of 1996, these individuals were contacted by telephone to determine their smoking status, their impression of the class, and their use of nicotine replacement products. Of the 160 individuals contacted, 33 (20 percent of those contacted and 15 percent of the total who attended the sessions) reported being smoke-free for six months. Another 15 percent reported that they were currently smoke-free, but had not been so for a full six months. Most felt that the session was worthwhile, regardless of current smoking status.

The most positive comments received were from those who expressed appreciation of the education and/or supplementary materials received and those who said they were grateful that they were able to quit. Some critical comments indicated a desire for more classes. Some respondents found the classes to be too large, wished the session would have been longer, or felt they had to wait too long for a new class to start.

A most surprising finding was the low rate of nicotine patch use. Although the offer of the free patch was meant to motivate smokers to consider the class, many used neither the patch nor the gum. Table 1 divides those who stopped smoking from present smokers, and reports the percentages in each category that used the patch or the gum. Overall, over 40 percent of attendees used neither product. Approximately 35 percent reported trying the patch; fewer than 20 percent tried the gum.

Providing nicotine dependence classes appears to be a valuable addition to patient services. The classes will continue as long as HealthPartners members continue to attend.

Research Team
GHF/HealthPartners-David H. Klevan MD, FACP; Brenda J. Talarico, BA; Jodi M. Lange; Mary Lou Beck, RN; Sharon J. Rolnick, PhD, MPH

The reason why our Smoke-Free (<6 months) and Present Day Smokers data
equal more than 100% is because some patients used both gum and patch products during their cessation attempt. We had 1 patient use both products in the Smoke-Free (<6 months) group, and 3 patients used both in the Present Day
Smokers group. We preferred not to ignore dual product use in our data set, so
that is why we included the disclaimers:
* 80% used the patch <4 weeks
** 1 patient used both patch and gum
*** 3 patients used both patch and gum
Nonetheless, this is the true data set.

return to top

One of the main objectives for all of the research conducted through Group Health Foundation is to use the research results to contribute to the HealthPartners mission-improving the health of our members and the community. Our largest project, IMPROVE (Improving Prevention through Organization, Vision and Empowerment), is demonstrating how this can be accomplished.

IMPROVE is designed to assess whether it is possible to increase the quantity and quality of clinical preventive services in primary care clinics by helping clinics develop better systems through Continuous Quality Improvement (CQI) techniques. Interestingly, this question is being tested by a unique and successful collaboration between HealthPartners and Blue Plus. The test involved recruiting 44 primary care clinics that contract with one or both HMOs and providing training, consultation, and networking to 22 of those clinics. The goal is to improve the way all the patients in these clinics receive services such as screening for breast and cervical cancer, screening and management of hypertension, hypercholesterolemia, tobacco use, and immunizations against influenza and pneumonia in the elderly.

This work with the 22 "intervention" clinics was completed in June of 1996. The clinics worked very hard on this project and nearly all said it was a very valuable experience. Although it was not necessary for the research, we offered to help the other 22 clinics and 17 accepted our offer. This provided the project an opportunity to test an approach that was improved based upon what we had learned. It also offered us an opportunity to pass the intervention on to the sponsoring HMOs.

Both Blue Plus and HealthPartners took advantage of this opportunity, and each have identified a physician and four consultants who will learn how to conduct this type of intervention over the next year. In addition, both HMOs have been so pleased with the process that they have recruited another 11 medical groups, most with many clinic sites. The responsibility for consulting with HMO-recruited clinics rests with HMO staff, while training for all sites is a shared, seamless operation.

After the research project is completed in June of 1997, the HMOs will continue to work with the clinics. Eventually, all members of both plans may benefit from these activities. In addition, the methods and knowledge will be made available to other HMOs nationwide.

Meanwhile, a separate research trial is being conducted in collaboration with the Minnesota Department of Health to apply the approach developed through the IMPROVE to improve the care of patients with diabetes and related chronic diseases. The original plan for Project IDEAL (Improving Diabetes Care through Empowerment, Active Collaboration and Leadership) involved implementing successful methods in the the HealthPartners system, as well as elsewhere in the state. The Centers for Disease Control and Prevention, which is funding this project, is eager to learn from the experience and to disseminate our methods to the rest of the nation.

Research Team
GHF/HealthPartners-Leif Solberg, MD; Andrew Nelson, MPH; Katie Giles, MBA; Ann Sisel, BA; Carol Westrum, MA; Carolyn Calomeni, RN, MPA; Charlie Straw, BS; Kathleen Conboy, BSN; Jerry Amundson, BA; Todd Gilmer, MA, Seongryeol Ryu, PhD

Blue Plus-Shirley Conn, RN, MSN; Sanne Magnan, MD

Mayo Foundation-Thomas E. Kottke, MD

Brekke Associates-Milo L. Brekke, PhD

MN Department of Human Services-Gestur Davidson, PhD

Funding
Agency for Health Care Policy and Research

return to top

Home / Investigators / Publications / Healthy Outcomes / Projects / Presentations / Research Administration / Internships / Career Opportunities / Maps

Copyright 2001 HealthPartners Research Foundation. All rights reserved.