Predictors of chronic kidney disease in long-term survivors of lung and heart-lung transplantation.

Renal insufficiency is common after non-renal organ transplants. The predictors of long-term renal outcomes are not well established. A total of 219 lung and heart-lung transplant recipients surviving more than 6 months after transplantation were studied to determine predictors of time to doubling of serum creatinine and end-stage kidney disease (ESKD) with death as a competing risk. Median follow-up was 79 months (range 9-222 months). Baseline estimated glomerular filtration rate (GFR) was 96.3+/-34.5 mL/min/1.73 m2. One hundred twenty-two recipients (55%) doubled their serum creatinine, 16 (7.3%) progressed to ESKD and 143 (65%) died. The majority of recipients who survived >6 years had a GFR<60 mL/min at both 1 and 7 years. Most of the loss of renal function occurred in the first year post-transplant. Older age at transplant, lower GFR at 1 month and cyclosporine use in the first 6 months predicted shorter time to doubling of serum creatinine when death was handled as a competing risk. Based on this prevalence data and using GFR decay and death as study endpoints, we offer sample size estimates for a prospective, interventional trial that is aimed at slowing or preventing the progression of kidney disease.

Reliability of a stool consistency classification system.

PURPOSE: Reliability and validity are important measures of the quality of a research or clinical instrument. This research determined the inter-rater and test-retest reliability of a stool consistency classification system and the agreement between the subjects' and expert investigators' classifications. METHODS: Two studies were conducted with adult volunteers. Stool consistency was described by words only and words + diagrams. To determine inter-rater reliability, subjects in each of 3 groups (20 nurses, 20 nursing students, and 20 lay persons) classified the consistency of 12 stool specimens. To determine test-retest reliability, 43 additional subjects classified the consistency of 9 stool specimens in 2 sessions. OUTCOME MEASURES: Outcome measures were reported per individual stool specimen and for all stool specimens. The outcome measures were the consistency classifications by the 3 groups of raters when the 2 types of descriptors (word only and word + diagram) were used. Other outcomes were the consistency classifications of the subjects on the 2 days of stool evaluation and the classifications of the subjects compared with those of the investigators. RESULTS: No significant difference was found among the stool consistency classifications among nurses, nursing students, or lay persons. Classifications were similar for 11 of 12 stool specimens when either word-only or word + diagram descriptions were used. No significant difference was found among the classifications between days 1 and 2. At least 75% of the subjects' classifications of stool consistency in both studies agreed highly with those of the investigators. CONCLUSIONS: The reliability and validity of the stool consistency classification system are good. The word-only descriptions of the consistency classifications appeared to be equally as effective as the word + diagram descriptions.

Supplementation with dietary fiber improves fecal incontinence.

BACKGROUND: Human studies have shown that dietary fiber affects stool composition and consistency. Because fecal incontinence has been shown to be exacerbated by liquid stools or diarrhea, management strategies that make stool consistency less loose or liquid may be useful. OBJECTIVE: To compare the effects of a fiber supplement containing psyllium, gum arabic, or a placebo in community-living adults who were incontinent of loose or liquid stools. Mechanisms underlying these effects (e.g., fermentation of the fibers and water-holding capacity of stools) were examined. METHODS: Thirty-nine persons with fecal incontinence of loose or liquid stools prospectively recorded diet intake and stool characteristics and collected their stools for 8 days prior to and at the end of a 31-day fiber supplementation period. During the fiber supplementation period, they ingested psyllium, gum arabic, or a placebo by random assignment. RESULTS: In the baseline period, the groups were comparable on all variables measured. In the fiber supplementation period, (a) the proportion of incontinent stools of the groups ingesting the fiber supplements was less than half that of the group ingesting the placebo, (b) the placebo group had the greatest percentage of stools that were loose/unformed or liquid, and (c) the psyllium group had the highest water-holding capacity of water-insoluble solids and total water-holding capacity. The supplements of dietary fiber appeared to be completely fermented by the subjects as indicated by nonsignificant differences in total fiber, short chain fatty acids and pH in stools among the groups in the baseline or fiber supplementation periods. CONCLUSIONS: Supplementation with dietary fiber from psyllium or gum arabic was associated with a decrease in the percentage of incontinent stools and an improvement of stool consistency. Improvements in fecal incontinence or stool consistency did not appear to be related to unfermented dietary fiber.

Use of complementary therapies, adherence, and quality of life in lung transplant recipients.

PURPOSES: The purpose of this study was to describe complementary and alternative medicine (CAM) use by lung transplant patients and to determine whether CAM users differ from nonusers with respect to health status, quality of life, or medical adherence. METHODS: A mailed survey seeking CAM, quality of life, and adherence information was sent to 145 lung transplant recipients, and 99 responded. RESULTS: The majority (88%) used at least 1 form of CAM (median, 2; range, 0-17). Prayer (68%), support groups (43%), and relaxation techniques (31%) were the most common. Only 44% of users reported discussing CAM with their providers. CAM users were adherent to their transplant regimen. Few differences were found between CAM users and nonusers. Education, high symptom burden, female sex, and depression symptoms were associated with various types of CAM use. CONCLUSION: Most lung transplant recipients are using CAM. Providers must explore potential for interaction or enhancement between CAM and standard therapy to optimize care.

Pre-transplant corticosteroid use and outcome in lung transplantation.

BACKGROUND: The early experience of lung transplantation was plagued with airway anastomotic complications. The use of corticosteroids in the pre-transplant period has been implicated as a major contributing factor in bronchial dehiscence, and many patients have been denied transplantation on the basis of corticosteroid use. We conducted the current study to assess the risks associated with pre-transplant corticosteroid use. METHODS: We analyzed records of 73 single- and bilateral-single lung transplant recipients who had chronic obstructive pulmonary disease or alpha(1)-antitrypsin deficiency as their underlying disease from 1986 to 1996. Twenty-six patients (steroid group) received daily corticosteroid therapy (prednisone, 1.5 to 40 mg/day) up to the time of transplantation, whereas 47 patients did not receive chronic corticosteroids and had no corticosteroid therapy within 3 months of transplantation (non-steroid group). RESULTS: The demographic profiles of the 2 groups were comparable. We noted no statistical significances in length of hospital stay, duration of intensive care, and post-operative pulmonary function. The rates of cytomegalovirus infection, acute rejection, bronchiolitis obliterans syndrome, and survival were also similar. The non-steroid group seemed to have a higher rate of bronchial stenosis at 3 years (29% vs 6%, p = 0.03). Bronchial dehiscence did not occur in either study group. CONCLUSIONS: Pre-transplant use of corticosteroids does not adversely affect outcome following lung transplantation.

Value-added outcomes: the use of advanced practice nurses in long-term care facilities.

The purpose of this study was to determine the effect on clinical outcomes for newly admitted nursing home residents when advanced practice gerontological nurses (APNs) worked with staff to implement scientifically based protocols for incontinence, pressure ulcers, depression, and aggressive behavior. Use of APNs in this manner differs from the usual way APNs have been used in nursing homes, in which their primary focus has been to augment the physician's role. The APN treatment was randomly assigned to two nursing homes and usual care was assigned to a third. Trajectories from admission to 6 months revealed that residents with APN input into their care (n = 86) experienced significantly greater improvement or less decline in incontinence, pressure ulcers, and aggressive behavior, and they had higher mean composite trajectory scores compared with residents receiving usual care (n = 111). Significantly less deterioration in affect was noted in cognitively impaired residents in the treatment group. Findings suggest that APNs can be effective links between current scientific knowledge about clinical problems and nursing home staff.

Development of health care delivery systems that are sensitive to the needs of stroke survivors and their caregivers.

The life-altering event of a stroke has long-term effects not only on stroke survivors but also on their caregivers, health care professionals, and health care delivery systems. The nurse administrator is faced with an obvious challenge to organize nursing systems to meet the multiple needs of the stroke survivor. The article presents data on the perceived needs of stroke survivors and their caregivers that provide direction and assistance to nursing administrators in organizing nursing services to address these perceived needs.

Development of a measure of resident satisfaction with the nursing home.

A satisfaction instrument specifically designed for use with nursing home residents, the Satisfaction with the Nursing Home Instrument (SNHI), was developed and tested with a sample of 110 nursing home residents from three proprietary facilities in Minnesota. As hypothesized, significant relationships were found between SNHI scores and measures of affect (negatively associated with depression and positively associated with morale), providing support for the construct validity of the scale. The lack of a significant relationship between SNHI scores and both age and mental status confirmed the predicted divergent validity of the instrument. The alpha coefficient for the 29-item scale was 0.81.

Fecal incontinence in hospitalized patients who are acutely ill.

BACKGROUND: Information about fecal incontinence experienced by patients in acute-care settings is lacking. The relationship of fecal incontinence to several well-known nosocomial or iatrogenic causes of diarrhea has not been determined. OBJECTIVES: To determine the cumulative incidence of fecal incontinence in hospitalized patients who are acutely ill, and to ascertain the relationship between fecal incontinence and stool consistency, and between diarrhea and two well-known nosocomial or iatrogenic etiologies of diarrhea: Clostridium difficile and tube feeding. The relationship of fecal incontinence and risk factors for diarrhea associated with C. difficile and tube feeding in hospitalized patients was examined. METHODS: Fecal incontinence, stool frequency and consistency, administration of tube feeding and medications, severity of illness, and nutritional data were prospectively recorded in 152 patients on acute or critical care units of a university-affiliated Veterans' Affairs Medical Center. Rectal swabs and stool specimens from patients were obtained weekly for C. difficile culture. C. difficile culture and cytotoxin assay were performed on diarrheal stools. HindIII restriction endonuclease analysis (REA) was used for typing of C. difficile isolates. RESULTS: In this study, 33% (50/152) of the patients had fecal incontinence. The proportion of total surveillance days with fecal incontinence in these patients was 0.50 +/- 0.06. A greater percentage of patients with diarrhea had fecal incontinence than patients without diarrhea (23/53 [43%] vs. 27/99 [27%]; p = 0.04). Incontinence was more frequent in patients with loose/liquid stool consistency than in patients with hard/soft stool consistency (48/50 [96%] vs. 71/100 [71%]; p < 0.001). The proportion of surveillance days with fecal incontinence was related to the proportion of surveillance days with diarrhea (r = 0.69; p < 0.001) and the proportion of surveillance days with loose/liquid stools (r = 0.64; p < 0.001). Multivariate risk factors for fecal incontinence were unformed/loose or liquid consistency of stool (RR = 11.1; 95% confidence interval [CI] = 2.2, 56.7), severity of illness (RR = 5.7; CI = 2.6, 12.3), and age (RR = 1.1; CI = 1, 1.1). CONCLUSIONS: Fecal incontinence is common in hospitalized patients who are acutely ill, but the condition was not associated with any specific cause of diarrhea. Because loose or liquid stool consistency is a risk factor for fecal incontinence, use of treatments that result in a more formed stool may be beneficial in managing fecal incontinence. However, treatments that slow intestinal transit should be avoided in patients with C. difficile-associated diarrhea.

Comparison of the nutritional composition of diets of persons with fecal incontinence and that of age- and gender-matched controls.

PURPOSE: In our clinical and research experience, persons with fecal incontinence anecdotally report altered diet intake to avoid incontinence. The purpose of this study was to compare the dietary intake of 39 persons with fecal incontinence living in the community with that of age- and gender-matched control subjects who had normal bowel function. The diets of both groups were compared with recommended dietary allowances (RDAs) for their constituent nutrients. METHODS: Subjects prospectively recorded the type, amount, and method of preparing all foods and beverages ingested for 8 consecutive days. Diet records were analyzed using the Nutritionist IV software program. MAIN OUTCOME MEASURES: The main outcome measures were the amounts and percentage of the RDAs of macronutrients and micronutrients in the subjects' diets. RESULTS: There were no significant differences in the intake of total kilocalories, protein, fat, dietary fiber, caffeine, or lactose by the fecal incontinence and control groups. The fecal incontinence group had a greater intake of carbohydrates, manganese, and vitamin B(1) compared with the control group. Diets of both groups exceeded 100% of the RDA for protein, phosphorus, iron, sodium, potassium, Vitamins B(1), B(2), B(3), B(12), and C and folate. Diets of both groups had less than 50% of the RDA for biotin, chromium, copper, and manganese but did not differ significantly. The percentages of the RDA for calcium and vitamin D were 84% +/- 6% and 56% +/- 8% for the fecal incontinence group and 90 +/- 8 (P =.6) and 69 +/- 11 for the control group (P =.4). CONCLUSIONS: The diets of persons with fecal incontinence were similar to those of control subjects with normal bowel function. Both the fecal incontinence and control groups may improve their nutritional patterns by lowering sodium and protein intake and increasing dietary fiber and monounsaturated fat intake. Calcium and vitamin D supplementation may improve dietary deficiencies and lower disease risks. Including a nutritional assessment and consultation in the care of persons with fecal incontinence to improve their general health and prevent disease is recommended, but consideration must be given to altered diet patterns perceived by the patient to prevent fecal incontinence.

Impact of transplantation on quality of life in patients with diabetes and renal dysfunction.

BACKGROUND: Simultaneous pancreas/kidney transplant (SPK) is an effective therapy that enables people with insulin-dependent diabetes mellitus (IDDM) and renal failure to maintain a more normal lifestyle, without the burdens of dialysis and insulin therapy. However, SPK has been viewed as a higher cost and higher risk procedure than kidney transplant, and it is unclear if SPK offers better health and quality of life (QOL) outcomes than insulin therapy plus kidney transplant alone (KTA). The purpose of this study is to determine which procedure affords better health and QOL outcomes. METHODS: This is a prospective observational study with assessments at pretransplant and 1 and 3 years posttransplant. Patients with IDDM and renal dysfunction who received either SPK or KTA from August 1990 to September 1993 at a university transplant center were enrolled. A convenience sample of patients with IDDM and complications not seeking transplants were enrolled during the same time interval. The main outcome measures were the SF-36 Short Form Health Survey and a Satisfaction with Diabetes Therapy Scale. RESULTS: Most health status and QOL measures improved from baseline values within each transplant group. After adjustment for diabetes severity and other baseline variables, year 3 follow-up scores of the SPK cohort were better than those of the KTA cohort for several SF-36 scales: physical functioning (P=0.038); bodily pain (P=0.047), general health (P=0.014), and the physical component summary (P=0.003). SPK recipients also reported greater satisfaction with diabetes therapy (P=0.014) and perceived more benefits to secondary complications. The KTA patients, however, had higher adjusted scores for the role-emotional subscale (P=.037) and the mental component summary (P=.037). By year 3, the SPK cohort is at the 30th and 51st percentiles of the general adult US population in self-reported physical and mental health; the KTA cohort is at the 10th and 73rd percentile. CONCLUSIONS: At follow-up, both SPK and KTA patients report better health and quality of life but SPK patients report greater improvements than KTA patients in physical health and in areas that are diabetes specific. Although the improved physical outcomes of SPK patients are consistent with perceived benefits to secondary complications, the mental health differences cannot be explained by the study data and warrant further study.

Peripheral blood allogeneic microchimerism in lung and cardiac allograft recipients.

We have been investigating two parameters, donor antigen-specific hyporeactivity and peripheral blood allogeneic microchimerism, to determine whether these parameters will predict a chronic rejection-free state and which recipients may be candidates for steroid withdrawal. We have identified donor antigen-specific hyporeactivity for 33% (16/48) of lung and 23% (11/47) of heart recipients. For both organ groups, the hyporeactive subgroup experienced a lower incidence of chronic rejection. The probability of donor antigen-specific hyporeactivity predicting a chronic rejection-free state is 100% for lung and 91% for heart recipients. We have identified peripheral blood allogeneic microchimerism for 77% (20/26) of lung and 36% (9/25) of heart recipients tested at 12-18 months posttransplant. Donor antigen-specific hyporeactivity correlates with a critical level of donor cells in lung recipients; the probability of high peripheral blood allogeneic microchimerism levels predicting a chronic rejection-free state in lung recipients is 100%. The results in heart recipients are not as clear with a short-, but not long-term, trend of higher chimerism levels correlating with the development of donor antigen-specific hyporeactivity. These results illustrate the usefulness of immmune parameters to predict long-term graft outcome in an organ-specific manner.

Adjuvant treatment of refractory lung transplant rejection with extracorporeal photopheresis.

BACKGROUND: Extracorporeal photopheresis is an immunomodulatory technique in which a patient's leukocytes are exposed to ultraviolet-A light after pretreatment with 8-methoxypsoralen (methoxsalen). There have been few reports describing the use of extracorporeal photopheresis in lung transplant recipients. METHODS: We reviewed our experience using extracorporeal photopheresis in 8 lung transplant recipients since 1992. All 8 patients had progressively decreasing graft function and 7 were in bronchiolitis obliterans syndrome grade 3 before the initiation of photopheresis. One patient had undergone a second transplant operation for obliterative bronchiolitis. Two patients had a pretransplantation diagnosis of chronic obstructive pulmonary disease, 1 alpha1-antitrypsin deficiency, 1 cystic fibrosis, 1 bronchiectasis, 1 idiopathic pulmonary fibrosis, and 2 primary pulmonary hypertension. Before refractory rejection developed, all patients had been treated with 3-drug immunosuppression and anti-T-cell therapy. The median time from transplantation to the start of extracorporeal photopheresis was 16.5 months and the median number of treatments was 6. RESULTS: The condition of 5 of 8 patients subjectively improved after extracorporeal photopheresis therapy. In these 5 patients photopheresis was associated with stabilization of the forced expiratory volume in 1 second. In 2 patients there was histologic reversal of rejection after photopheresis. With a median follow-up of 36 months, 7 patients are alive and well. Three patients required retransplantation at a median of 21 months after completion of the treatments. Four patients have remained in stable condition after photopheresis. There were no complications related to extracorporeal photopheresis. CONCLUSION: We believe that this treatment is a safe option for patients with refractory lung allograft rejection when increased immunosuppression is contraindicated or ineffective.

Factors associated with clients not keeping their initial home health visit.

Public health nursing home visits are one strategy used by health care providers and communities to improve the health of pregnant women and families with young children. This study investigates the influence of client characteristics, nurse characteristics, and nurse-client interaction on clients keeping or not keeping their initial maternal and child health promotion home visit appointments with a public health nurse. Results of the study suggest the hiring decisions for nurse home visitors should give priority to the nurse's experience and education, as well as his/her practice style of handling maternal/child health visits.

Comparison of subjective classification of stool consistency and stool water content.

OBJECTIVE: The purpose of this study was to compare the stool consistency categorizations made by 39 adults with fecal incontinence and the percentage of water in their stools determined by lyophilization. METHODS: Subjects collected all stools daily for 8 days during a baseline period and at the end of a fiber treatment period. Stool consistency was recorded as hard and formed, soft but formed, loose and unformed, or liquid. Aliquots of the stools were lyophilized to constant weight. MAIN OUTCOME MEASURES: The main outcome measures were the percentage of stool water among stools in each consistency category and the correlation between subjects' stool consistency categorizations and the percentage of stool water. RESULTS: The subjects were 8 men and 21 women, ranging in age from 30 to 89 years, who were participating in a study of the effectiveness of dietary fiber for treating fecal incontinence. A total of 1023 stool samples were analyzed. Significant differences in the mean percentage of water were found among the 4 stool consistency categories (hard and formed = 68% +/- 0.9%, soft but formed = 74% +/- 0.3%, loose and unformed = 80% +/- 0.4%, and liquid = 85% +/- 0.3%; P < .001). Ninety-six percent of the stools had a percentage of water within 2 SDs of the mean percentage of water of other stools in their consistency category. CONCLUSION: This classification system of stool consistency is a valid and practical measure for clinical studies. It may be useful for clinicians and patients to evaluate outcomes of treatments directed at improving stool consistency.

A theory of planned behavior-based structural model for breast-feeding.

BACKGROUND: Breast-feeding is the recommended method of infant feeding because it is clearly associated with health benefits for infants and their mothers. Yet, many women who initiate breast-feeding fail to meet their own personal goals or recommended standards for duration of breast-feeding. OBJECTIVE: To refine a Theory of Planned Behavior (TPB)-based structural model for explaining variability in breast-feeding intention and duration. METHOD: The study design was prospective, multicorrelational, and longitudinal. Out of the total sample of 635 women, 602 mothers of healthy, full-term infants provided complete datasets over the entire course of their breast-feeding experience and these datasets were used in the modeling analyses. Simultaneous multisample analysis of covariance structures was used to develop the model. RESULTS: The resulting TPB for Breast-Feeding (TPB-BrF) describes the rational, motivational processes of the original TPB, but reconfigures the relationships among them, for homemakers (TPB-BrF/H), women employed half-time or less (TPB-BrF/EL), and women employed more than half-time (TPB-BrF/EM). Mothers' early postpartum ratings of adequacy of milk supply and stimulus conditions of maternal education and breast-feeding knowledge were included in the TPB-BrF to better explain breast-feeding outcomes. Model complexity increased with employment effort. CONCLUSION: The TPB-BrF is a comprehensive, theoretically based, empirically verified model that can serve as a useful heuristic for understanding the personal motivational components of breast-feeding behavior.

Acquisition of Clostridium difficile and Clostridium difficile-associated diarrhea in hospitalized patients receiving tube feeding.

BACKGROUND: Clostridium difficile is the most common infectious cause of nosocomial diarrhea, but its role in diarrhea associated with tube feeding has not been rigorously investigated. OBJECTIVE: To determine the incidence of C. difficile acquisition and C. difficile-associated diarrhea in tube-fed and non-tube-fed patients. DESIGN: Prospective cohort study. SETTING: A university-affiliated Veterans Affairs Medical Center. PATIENTS: 76 consecutive hospitalized, tube-fed patients and 76 hospitalized, non-tube-fed patients. The two cohorts were matched for age, unit location, duration of hospitalization before surveillance, and severity of illness. MEASUREMENTS: Incidence of C. difficile acquisition, incidence of C. difficile-associated diarrhea, and C. difficile restriction endonuclease analysis typing results. RESULTS: More tube-fed patients than non-tube-fed patients acquired C. difficile (15 of 76 patients [20%] compared with 6 of 76 patients [8%]; P=0.03) and developed C. difficile-associated diarrhea (7 of 76 patients [9%] compared with 1 of 76 patients [1%]; P=0.03). The mean proportion (+/-SD) of surveillance days with diarrhea was greater for tube-fed patients after the development of C. difficile-associated diarrhea than for tube-fed patients without this diarrhea (0.68+/-0.4 compared with 0.22+/-0.2 [95% CI for the mean difference, 0.08 to 0.84]). Postpyloric tube feeding (odds ratio, 3.14 [CI, 1.008 to 9.77]) and duration of surveillance (odds ratio, 1.08 [CI, 1.0009 to 1.16]) were risk factors for the acquisition of C. difficile. Nineteen restriction endonuclease analysis types of C. difficile were identified from 20 patients. CONCLUSIONS: Hospitalized, tube-fed patients, especially those receiving postpyloric tube feeding, are at greater risk for the acquisition of C. difficile and the development of C. difficile-associated diarrhea than are hospitalized, non-tube-fed patients. Clinicians should test for C. difficile in tube-fed patients with diarrhea.

Outcomes of macrosomic infants in a nurse-midwifery service.

The purpose of this study was to compare outcomes of term infants of average birth weight with outcomes of large infants in a nurse-midwifery service. A retrospective study design was used. Data were retrieved from a computer data base that contained information from a data form routinely completed for all births. Subjects were women cared for by the nurse-midwives including those for whom the birth was conducted by a physician. The final n study population was 2,228; 322 (14.5%) of the infants weighed 4,000 g or more. Women who delivered large infants had a significantly higher prepregnant body mass index and pregnancy weight gain. Shoulder dystocia occurred more often in large infants; however, newborn intensive care unit admission rates did not differ between the average birth weight and the large infants. Apgar scores at 1 and 5 minutes were significantly lower for infants weighing > or = 4,500 g compared to those with birth weights of 2,500-3,999 g and those 4,000-4,449 g; however, these differences were not clinically significant. A trend for fewer occurrences of shoulder dystocia in the side-lying birth position was observed. Logistic regression predicting poor Apgar scores (< 7) showed parity as a protective factor and increased gestational age and higher maternal body mass index as predictive of low Apgar scores. Large infants had birth outcomes comparable to those reported by others in the medical literature, suggesting that nurse-midwifery management, including consultation with physician colleagues, can be appropriate and safe.

Factors associated with very early weaning among primiparas intending to breastfeed.

OBJECTIVES: The major objective of this study was to identify predictor variables that accurately differentiated breastfeeding women who weaned during the first 4 weeks, those who weaned between 5 and 26 weeks, and those who weaned after 26 weeks. Predictors were demographic variables, Theory of Planned Behavior (TPB) variables, breastfeeding knowledge, and difficulties experienced during the first month. METHODS: Primiparas who delivered healthy infants in an urban midwestern hospital provided initial data prior to discharge. Follow-up occurred at 1, 3, 6, 9, and 12 months. Following appropriate bivariate analyses, polychotomous logistic regression was used to determine predictors of weaning group. Linear multiple regression was used to predict intended duration. RESULTS: Most of the 84 women who weaned very early had intended to breastfeed considerably longer. According to the multivariate analysis, women who weaned earlier were younger, had completed fewer years of education, had a more positive bottle-feeding attitude and a less positive breastfeeding attitude, intended to breastfeed less time, had lower knowledge scores, had higher perceived insufficient milk scores, and planned to work outside the home. Variables postulated by the TPB to be direct predictors of intention explained 36% of the variance in intended duration. CONCLUSIONS: Women at risk for early weaning can be identified with reasonable accuracy using a TPB-based conceptual framework expanded to include breastfeeding specific variables. Casefinding using empirically derived screening methods and careful postpartum follow-up, along with professional intervention, should be used to avert unintended early weaning.