Evaluating the process of tailoring the global HIV/AIDS practice guidelines for use in developing countries.

OBJECTIVE: To describe a systematic procedure for adapting, or 'tailoring' the World Health Organisation's 'global guidelines for the management of HIV/AIDS in adults and children' for use in two developing countries: Malawi and Barbados. DESIGN: In order for these guidelines to achieve reproducibility, clinical flexibility, and clinical applicability, a systematic procedure is needed to tailor the guidelines to the local practice conditions of specific settings. METHODS: A group of local experts in each country used a nominal group process to modify the global program on AIDS (GPA) guidelines for local use. Semantic analysis techniques, known as clinical algorithm nosology (CAN), were used to compare the two modified guidelines with the global ones to determine the extent and type of differences between sets of guidelines. RESULTS: Standard, locally-tailored algorithm map guidelines (AMG) were developed within 4 months. CAN semantic analysis showed that guideline structure was maintained; 572/858 (66.6%) decision nodes were found to be the same in the GPA/Malawi, GPA/Barbados and Malawi/Barbados comparisons. However, different guideline versions managed patients quite differently, as evidenced by clinical algorithm patient abstraction (CAPA) scores of between 0 and 8.46 (0 = different; 8 = similar; 10 = identical). Analysis of the 197 specific differences found in these abstractions showed that 83% were in approaches to diagnosis and therapy, while the remaining 17% related to disease prevalence. CONCLUSIONS: Standard techniques involving consensus used to develop clinical guidelines can also be employed to tailor these guidelines to local settings. Semantic analysis shows that the tailoring preserves structure but may involve significant modification to the processes of clinical care that could in turn affect care outcomes.

Text-to-algorithm conversion to facilitate comparison of competing clinical guidelines.

This study aimed to test the usefulness and reliability of text-to-algorithm conversion in comparing competing clinical guidelines and defining their differences. Two pairs of competing guidelines for measles immunization, published in 1989 and 1994, were analyzed and compared. Five categories of differences were detected: differences in recommendations, excluded elements, logical inconsistencies, nonspecific phrases, and approaches to contraindications. On a scale of 0-10 (where identical=10), the overall comparison scores were 6.01 for the guidelines published in 1989 and 5.54 for the guidelines published in 1994. Text-to-algorithm conversions performed by three different persons on the 1989 guidelines were compared and found similar. Text-to-algorithm conversion is an important step in facilitating comparison of competing guidelines. It has the potential to assist in making rational and systematic choices between competing guidelines before actual field testing takes place. Physicians can use it to analyze and to learn a prose clinical guideline, to critique existing guidelines, and to simulate hypothetical patients for teaching and evaluating clinical management.

Is consensus reproducible? A study of an algorithmic guidelines development process.

The authors evaluated the reproducibility of a clinical algorithm consensus development process across three different physician panels at a health maintenance organization. Physician groups were composed of primary care internists, who were provided with identical selections from the medical literature and first-draft "seed" algorithms on the management of two common clinical problems: acute sinusitis and dyspepsia. Each panel used nominal group process and a modified Delphi method to create final algorithm drafts. To compare the clinical logic in the final algorithms, the authors applied a new qualitative and quantitative comparison method, the Clinical Algorithm Patient Abstraction (CAPA). Dyspepsia algorithms from all physician groups recommended empiric anti-acid therapy for most patients, favored endoscopy over barium swallow, and had very similar indications for endoscopy. The average CAPA comparison score among final physician algorithms was 6.1 on a scale of 0 (different) to 10 (identical). Sinusitis algorithms from all groups proposed empiric antibiotic therapy for most patients. Indications for sinus radiographs were similar between two algorithms (CAPA = 4.9), but differed significantly in the third, resulting in lower CAPA scores (average CAPA = 1.9, P < 0.03). The clinical similarity of the algorithms produced by these physician panels suggests a high level of reproducibility in this consensus-driven algorithm development process. However, the difference among the sinusitis algorithms suggests that physician consensus groups using a consensus process that a health maintenance organization can do with limited resources will produce some guidelines that vary due to differences in interpretation of evidence and physician experience.

Physician use of a computerized medical record system during the patient encounter: a descriptive study.

The computerized medical record program, CLINIC, specifically designed at Ben-Gurion University of the Negev to be used on-line in a primary care clinic during a patient encounter by healthcare professionals, has been in use for over 6 years in two community clinics. In order to analyse the effectiveness and use of CLINIC, four physicians were video recorded during working sessions before and after the introduction of CLINIC. While using CLINIC did not change the total mean encounter time, the lengths of the encounter components and record use did change. The physicians' work style changed from a 'conversational pattern' (continuous data recording) to a 'blocked pattern' (data entry at intervals).

The clinical algorithm nosology: a method for comparing algorithmic guidelines.

Concern regarding the cost and quality of medical care has led to a proliferation of competing clinical practice guidelines. No technique has been described for determining objectively the degree of similarity between alternative guidelines for the same clinical problem. The authors describe the development of the Clinical Algorithm Nosology (CAN), a new method to compare one form of guideline: the clinical algorithm. The CAN measures overall design complexity independent of algorithm content, qualitatively describes the clinical differences between two alternative algorithms, and then scores the degree of similarity between them. CAN algorithm design-complexity scores correlated highly with clinicians' estimates of complexity on an ordinal scale (r = 0.86). Five pairs of clinical algorithms addressing three topics (gallstone lithotripsy, thyroid nodule, and sinusitis) were selected for interrater reliability testing of the CAN clinical-similarity scoring system. Raters categorized the similarity of algorithm pathways in alternative algorithms as "identical," "similar," or "different." Interrater agreement was achieved on 85/109 scores (80%), weighted kappa statistic, k = 0.73. It is concluded that the CAN is a valid method for determining the structural complexity of clinical algorithms, and a reliable method for describing differences and scoring the similarity between algorithms for the same clinical problem. In the future, the CAN may serve to evaluate the reliability of algorithm development programs, and to support providers and purchasers in choosing among alternative clinical guidelines.

Computerized algorithms and pediatricians' management of common problems in a community clinic.

In 1987, a microcomputer clinical algorithm (CA) system for constructing and using CAs for patient care was designed and implemented for six common primary care pediatrics problems. Six community clinic pediatricians agreed to use the system for several months. Length of patient's visit, completeness of data collection, antibiotic use, and appropriateness of clinical plan were measured before the computers were introduced (without CAs) and after the computers were introduced (both with and without CAs). All performance measures improved after the introduction of CAs. However, CA implementation had to be discontinued after five weeks because the CAs were too tedious for the physicians to follow during routine care. The authors conclude that CAs cannot be successfully sustained with physicians for common problems, even though their design and use can significantly improve the process of care.

Cryptosporidium in Bedouin and Jewish infants and children in southern Israel.

We conducted a 1-year prospective study in two clinics and a hospital in the Negev region of southern Israel to determine the epidemiological and clinical patterns of Cryptosporidium diarrhea among Jewish and Bedouin infants and children living in the same geographical area. A total of 612 episodes were studied: 398 in Bedouins and 214 in Jews, of which 449 occurred in patients with diarrhea and 164 in controls. Cryptosporidium was detected in 13 of 382 patients (3.4%) with diarrhea and in 1 of 138 controls (0.7%) (P = 0.078). In 5 of 13 Cryptosporidium-positive patients (38%) another pathogen was detected. No significant difference in Cryptosporidium detection rates was observed between Jews and Bedouins or between hospitalized or nonhospitalized patients. The frequency of Cryptosporidium detection did not differ significantly when three age-groups were compared (less than 6 months old, 7-12 and 13-36 months old). The rate of Cryptosporidium detection was similar among malnourished and well-nourished patients, as determined by weight-for-height percentiles. Cryptosporidium was detected more frequently during the summer months (8.3%) than during the rest of the year (1.2%) (P less than 0.001). Patients with Cryptosporidium diarrhea did not differ clinically from patients with other causes of diarrhea. However, they were characterized by the absence of fecal leukocytes. Cryptosporidium is not a rare cause of diarrhea in southern Israel. It is more prevalent during the hot and dry season and can be detected in a relatively high prevalence among very young infants. Its clinical features are indistinguishable from those of patients with non-cryptosporidial diarrhea.

A computerized medical record with direct data entry for community clinics in Israel.

CLINIC is a computerized medical record system currently being used in two primary care clinics in Israel. Clinic features direct coded data-entry by the medical personnel via a system based on categories of problems and complaints with common signs and symptoms.

Rheumatology algorithms for primary care physicians.

Primary care physicians were trained on three rheumatology topics to assess the effectiveness of an educational strategy for continuing medical education. Algorithm training was shown to be at least as effective as that based on standard prose monographs. Both training groups improved their knowledge of patient management skills but there were no statistically significant differences between groups in the amount learned. When algorithms were used to design text materials, the designed texts required less study time than did the annotated clinical algorithms alone. That difference was significant for the shoulder pain materials (P less than 0.05) but not for the osteoporosis materials. The ratio of knowledge gained to study time was significantly higher for the algorithm group on the low back pain topic (P less than 0.05) but not for the other topics. Taped interview problems tests were studied as a method for assessing patient management skills related to problem-specific indicator conditions and were found to produce interrater reliability greater than 0.80 on five of the six tests.

Rotavirus diarrhea in Jewish and Bedouin children in the Negev region of Israel: epidemiology, clinical aspects and possible role of malnutrition in severity of illness.

We conducted a 1-year prospective study in the Negev region of southern Israel to determine the epidemiologic and clinical patterns of rotavirus diarrhea. A total of 605 patients were studied, 392 Bedouins and 213 Jews, 441 of whom had diarrhea (449 episodes) and 164 did not. Rotavirus was the most common organism detected in children with diarrhea (63 of 444; 14%) but was rarely found in controls (3 of 163; 2%) (P less than 0.001). In 22% (12 of 54) of the rotavirus-positive patients, at least one other organism was also detected. The rate of rotavirus detection decreased as age increased, from 18% in the first year to 8% in the third year of life. Hospitalization with rotavirus diarrhea occurred more frequently in the summer. However, during winter, when diarrhea was less prevalent in the community, the proportion of cases associated with rotavirus was higher. Compared with controls, malnourished children were more likely to be hospitalized. However, rotavirus was detected in similar proportions among well-nourished and malnourished cases with diarrhea. The most prevalent rotavirus serotype was type 1 (in 69%), followed by types 4 and 2 (18 and 13%, respectively). We estimated that during the study period, approximately 2% of all Bedouin infants vs. only 0.2% of Jewish infants were hospitalized with rotavirus disease in their first year of life. Clinical signs and symptoms and stool appearance were not useful in predicting rotavirus detection. Malnutrition seems to be an important indicator of disease severity, which may explain why the toll of rotavirus-associated morbidity and mortality is particularly high among children in developing countries.

Clinical practice guidelines at an HMO: development and implementation in a quality improvement model.

Harvard Community Health Plan (HCHP) is adapting to clinical medicine the managerial principles and methods of quality improvement theory that were originally developed and successfully applied in industrial settings. An essential step in applying the quality improvement cycle to clinical medicine is the setting of standards or specifications for clinical care. HCHP has chosen to focus its standard-setting efforts on the development of clinical algorithms, which provide an excellent basis for specifying and communicating optimal care processes and for evaluating actual clinical care. When implemented effectively, clinical algorithms may improve quality and decrease costs by guiding clinicians toward more standardized, high-quality, cost-effective clinical strategies and by facilitating more valid measurement of clinical process and outcomes. This article describes the evolution, structure, methods, and future agenda of the Algorithm Based Clinical Quality Improvement Process (ABCQIP) at HCHP.

Clinical algorithms teach pediatric decisionmaking more effectively than prose.

Despite the rapidly increasing volume of medical literature, little attention has been paid to the appropriate printed format for teaching clinical content. This study attempted to determine whether a clinical algorithm (CA) or prose is more effective for teaching clinical decisionmaking. Clerkship students, preclerkship students, and pediatric house officers in five medical centers in the USA and Israel were presented with clinical algorithms and prose describing management of fever in a child under 2 years of age, and management of meningitis in children. Knowledge of decisionmaking was measured before and after learning, using audio-taped clinical problems and learning time as measured. It was concluded that CAs are more effective and more efficient than prose for teaching clinical decisionmaking. When writing about clinical decisionmaking, the use of CAs should always be considered, especially if a series of interdependent decisions is being described.