Progress towards the year 2000: assessing the health status of minorities in Texas

In 1991, the Centers for Disease Control and Prevention released a consensus set of 18 health status indicators to help communities assess their general health status and focus local, state, and national efforts on tracking the year 2000 objectives. States and local communities were encouraged also to analyze the indicators for each major racial and ethnic population in their jurisdictions. This article presents an update of trends in Texas for the health status indicators for the total population, along with comparisons by race and Hispanic origin for the most recent data. Analysis of these data reveals that the health status for the total population is improving as we approach the year 2000. Yet despite these overall gains, significant disparities exist between the total population and the racial and ethnic minorities in Texas. Further efforts are needed to reduce these health disparities.

A double-blind comparison of the efficacy of two dose regimens of oral granisetron in preventing acute emesis in patients receiving moderately emetogenic chemotherapy

BACKGROUND

The purpose of this study was to define an optimal administration schedule of granisetron for patients receiving moderately emetogenic chemotherapy by comparing the antiemetic efficacy and safety of 2 mg of the drug administrated orally.

METHODS

In this double-blind, randomized, parallel study, 2-dose regimens of oral granisetron were evaluated in 697 adult cancer patients. Patients were stratified by gender and randomized to receive 2 mg oral granisetron, either as a divided dose given 1 hour prior to chemotherapy and 12 hours after the start of chemotherapy, or as a single dose 1 hour prior to chemotherapy at Cycle 1. The primary efficacy endpoints assessed were the percentage of patients with complete response (no nausea, no emesis, and no additional antiemetic medication during the 24-hour post-chemotherapy interval) and the incidence of emesis and nausea. Following completion of Cycle 1, patients were given the opportunity to receive open-label granisetron (2 mg once daily) on the first day of each remaining cycle of chemotherapy.

RESULTS

No statistically significant differences in any of the endpoints were observed between the two treatment groups. Approximately 50% of patients in both treatment groups achieved complete response. The proportion of patients with no episodes of emesis occurred with similar frequency in the two treatment groups. Approximately 52% of patients in either treatment group were free of nausea during the postchemotherapy period. There was no difference between treatment groups regarding the use of antiemetic rescue medication. Finally, the incidence of adverse experiences was similar for both treatment groups.

CONCLUSIONS

Both dose regimens of oral granisetron were similarly effective in controlling nausea and vomiting in the 24-hour interval following chemotherapy. Granisetron was well tolerated with few adverse events attributable to the study drug.

Efficacy and safety of oral granisetron versus oral prochlorperazine in preventing nausea and emesis in patients receiving moderately emetogenic chemotherapy

PURPOSE

To compare the efficacy and safety of oral granisetron hydrochloride tablets with that of oral prochlorperazine sustained-release capsules in preventing nausea and emesis induced by moderately emetogenic chemotherapeutic agents.

PATIENTS AND METHODS

In this multicenter, double-blind, randomized, parallel group study, oral granisetron and oral prochlorperazine were compared in 230 chemotherapy-naive, adult cancer patients who received moderately emetogenic chemotherapy. Patients were stratified by gender and randomized to receive either 1.0 mg oral granisetron HCI twice a day for 7 days, or 10 mg oral prochlorperazine sustained-release capsules twice a day for 7 days. The first dose was given 1 hour before initiation of chemotherapy and the second dose 12 hours after the first dose. Patients were evaluated for emetic episodes, extent of nausea, and adverse events for 7 days after the start of chemotherapy. Primary efficacy parameters were complete response (no emetic episodes, no greater than mild nausea, no antiemetic rescue) and total control (no emetic episodes, no nausea, no antiemetic rescue) in the 24 hours after the start of chemotherapy.

RESULTS

Granisetron was significantly more effective than prochlorperazine in achieving a complete response (74% vs. 41%, respectively) and total control of nausea and vomiting (58% vs. 33%, respectively) at the 24-hour assessment. Complete response at 24 hours was significantly higher in the granisetron-treated patients than in prochlorperazine-treated patients. In women, granisetron showed a complete response rate of 69% versus 38% with prochlorperazine; in men, granisetron showed a complete response rate of 92% versus 61% with prochlorperazine. Both regimens were well tolerated, with headache (36% for granisetron, 29% for prochlorperazine) and constipation (31% for granisetron, 6% for prochlorperazine) the most common adverse events.

CONCLUSIONS

: Oral granisetron 1 mg twice a day was significantly more effective than oral prochlorperazine sustained release capsules 10 mg twice a day in complete response and total control of nausea and vomiting at 24 hours after chemotherapy. Both agents were well tolerated.

Identification of suspected tuberculosis patients based on natural language processing of chest radiograph reports

Identification of eligible patients from electronically available patient data is a key difficulty in computerizing clinical practice guidelines because a large amount of the relevant data is stored as free text. We have been using MedLEE (Medical Language Extraction and Encoding System), a natural language processing system, to encode the clinical information in all chest radiograph and mammogram reports. This paper describes a retrospective study to determine if MedLEE can identify patients at risk for having tuberculosis (TB) based on their admission chest radiographs. Reports of 171 adult inpatients with culture-positive TB during 1992 and 1993 were manually coded (by a TB specialist) using seven terms suggestive of TB, and were also encoded by MedLEE. Using manual coding as the gold standard, MedLEE agreed on the classification of 152/171 (88.9%) reports--129/142 (90.8%) suspicious for TB and 23/29 (79.3%) not suspicious for TB; and 1072/1197 (89.6%) terms indicative of TB. Analysis showed that most of the discrepancies were caused by MedLEE not finding the location of the infiltrate. By ignoring the location of the infiltrate, the agreement became 157/171 (91.8%) reports and 946/1026 (92.2%) terms. Thus, natural language processing offers a practical alternative for using free-text reports to determine patient eligibility for computerized clinical practice guidelines.

The reinforcement of cancellous bone screws with calcium phosphate cement

The ability of calcium phosphate cement (CPC) to reinforce cancellous screws placed in previously stripped holes was studied in vitro. The distal end of canine femurs were harvested. A total of 15 screws were placed in six femurs. The pullout strength (failure force), failure displacement, stiffness, and energy absorbed were determined for the screws in the intact cancellous bone. Next, these stripped screw holes were packed with CPC. The pullout test was repeated, and the results were compared using a paired, Student's t test. We found that the CPC was able to reinforce the previously stripped holes and significantly increase the pullout strength (1,159 +/- 278 N versus 678 +/- 297 N) and the stiffness (1,990 +/- 569 N/mm versus 1,519 +/- 609 N/mm) of the constructs, as well as the energy absorbed by the constructs until failure (467 +/- 180 N.mm versus 278 +/- 140 N.mm). There was no difference in the failure displacement (0.94 +/- 0.23 versus 0.85 +/- 0.51 mm). This study documents the ability of CPC to acutely reinforce cancellous bone screws in a region with no or poor-quality cancellous bone.

Integrating data from natural language processing into a clinical information system

Demographic data extracted from discharge summaries by natural language processing was compared to data gathered by a conventional hospital admitting system. Discrepancies in data were noted in names, age, sex, race, and ethnicity. Some differences are attributable to errors in collection: interaction with patient, dictation, transcription, and data entry. Very few differences were due to errors in natural language processing. Other differences can be used to critique existing data, or to enhance data with more detailed information. Discrepancies in data as elementary as patient demographics raise the issue of resolving conflicts when neither source of data is known to be more reliable. Clinical repositories can represent conflicting data from multiple sources, but clinical information systems must bear the cost of increased complexity in the application programs that will use the data.

Unlocking clinical data from narrative reports: a study of natural language processing

OBJECTIVE

To evaluate the automated detection of clinical conditions described in narrative reports.

DESIGN

Automated methods and human experts detected the presence or absence of six clinical conditions in 200 admission chest radiograph reports.

STUDY SUBJECTS

A computerized, general-purpose natural language processor; 6 internists; 6 radiologists; 6 lay persons; and 3 other computer methods.

MAIN OUTCOME MEASURES

Intersubject disagreement was quantified by "distance" (the average number of clinical conditions per report on which two subjects disagreed) and by sensitivity and specificity with respect to the physicians.

RESULTS

Using a majority vote, physicians detected 101 conditions in the 200 reports (0.51 per report); the most common condition was acute bacterial pneumonia (prevalence, 0.14), and the least common was chronic obstructive pulmonary disease (prevalence, 0.03). Pairs of physicians disagreed on the presence of at least 1 condition for an average of 20% of reports. The average intersubject distance among physicians was 0.24 (95% Cl, 0.19 to 0.29) out of a maximum possible distance of 6. No physician had a significantly greater distance than the average. The average distance of the natural language processor from the physicians was 0.26 (Cl, 0.21 to 0.32; not significantly greater than the average among physicians). Lay persons and alternative computer methods had significantly greater distance from the physicians (all > 0.5). The natural language processor had a sensitivity of 81% (Cl, 73% to 87%) and a specificity of 98% (Cl, 97% to 99%); physicians had an average sensitivity of 85% and an average specificity of 98%.

CONCLUSIONS

Physicians disagreed on the interpretation of narrative reports, but this was not caused by outlier physicians or a consistent difference in the way internists and radiologists read reports. The natural language processor was not distinguishable from the physicians and was superior to all other comparison subjects. Although the domain of this study was restricted (six clinical conditions in chest radiographs), natural language processing seems to have the potential to extract clinical information from narrative reports in a manner that will support automated decision-support and clinical research.

Comparative clinical trial of granisetron and ondansetron in the prophylaxis of cisplatin-induced emesis. The Granisetron Study Group

PURPOSE

To compare the efficacy and safety of granisetron and ondansetron, serotonin (5-HT3) receptor antagonists shown to be effective in the prevention of chemotherapy-induced emesis.

PATIENTS AND METHODS

In a double-blind, randomized, stratified, parallel-group study, the efficacy and safety of granisetron and ondansetron were compared in 987 chemotherapy-naive patients who received cisplatin in doses > or = 60 mg/m2. Granisetron was administered as a single dose of 10 or 40 micrograms/kg before the start of chemotherapy. Ondansetron was administered in doses of 0.15 mg/kg before and 4 and 8 hours after the start of chemotherapy. The three treatment groups were well-matched with respect to demographic characteristics and the dose of cisplatin administered.

RESULTS

For all evaluations, single doses of granisetron 10 or 40 micrograms/kg were as effective as three 0.15-mg/kg doses of ondansetron. Total control (no vomiting, no retching, no nausea, and no use of rescue) was attained by 38%, 41%, and 39% of all patients who received granisetron 10 microgram/kg, granisetron 40 micrograms/kg, and ondansetron, respectively. No vomiting or retching and no use of rescue antiemetics were reported in 47%, 48%, and 51% of patients who received granisetron 10 micrograms/kg, granisetron 40 micrograms/kg, and ondansetron, respectively; no nausea and no use of rescue antiemetics were reported in 39%, 42%, and 40% of patients, respectively.

CONCLUSION

All three treatment regimens were well-tolerated. The results of this study indicate that a single dose of granisetron 10 or 40 micrograms/kg is as effective as three doses of ondansetron 0.15 mg/kg in the prevention of nausea and vomiting induced by cisplatin chemotherapy.

Assignment of the human p27Kip1 gene to 12p13 and its analysis in leukemias

The p27Kip1 (p27) gene encodes an inducible inhibitor of cyclin-dependent kinase activity. Using a murine p27 cDNA as probe, we obtained a human cDNA clone and subsequently used it to isolate a genomic clone of this gene. The coding region of the human p27 gene was contained in two exons. Both the amino acid sequence and intron-exon organization of p27 were similar to those previously found for the related cyclin-dependent kinase inhibitor p21Waf1 (p21). The p27 gene was localized to chromosome band 12p13 by a combination of somatic cell hybrid and fluorescence in situ hybridization analyses. The p27 gene product is thought to control the leukocyte cell cycle and the 12p13 chromosomal band is known to be deleted in leukemias, suggesting that the p27 gene may act as a tumor suppressor gene in leukemias. Although p27 was found to reside in the minimal region of chromosomal loss in hematological malignancies, no mutations of p27 were observed in leukemia samples. Haploinsufficiency of p27 may confer a growth advantage to leukemia cells.

Comparison of case-finding methodologies for endometritis after cesarean section

BACKGROUND

Endometritis is a possible complication of delivery among patients undergoing cesarean section, resulting in increased costs and patient morbidity. However, traditional case-finding methods for endometritis may not identify most cases. We compared various case-finding methods with a reference method to determine a simple and accurate method for collecting data on endometritis after cesarean section.

METHODS

We reviewed charts of all patients undergoing cesarean section (N = 167) during March 1 through July 31, 1991. These data were compared with study case-finding methods that used microbiology data, infection report forms from nursing, and computerized reports linking patients undergoing cesarean section with intravenous antibiotic use data and admission and discharge diagnoses.

RESULTS

Each case-finding method was compared separately with the reference method ("gold standard"), which was designed to capture all cases among the patients in the study population (N = 145). This review yielded nine cases of endometritis (infection rate of 5.4/100 procedures). The computerized report method linking patients who underwent cesarean section with antibiotic use had a positive predictive value of 0.53. Methods that used microbiology data and nursing report forms had lower positive predictive values of 0.18 and 0.20, respectively.

CONCLUSIONS

In our institution, case finding for postcesarean endometritis by means of a computerized report linking patients undergoing cesarean section with i.v. antibiotic use data and admission and discharge diagnoses is the most effective method of detecting postcesarean endometritis. It also represents the most efficient use of the infection control department's resources.

Dinucleotide repeats in the human surfactant protein-B gene and respiratory-distress syndrome

Pulmonary surfactant, a lipoprotein complex, is essential for normal lung function, and deficiency of surfactant can result in respiratory-distress syndrome (RDS) in the prematurely born infant. Some studies have pointed towards a genetic contribution to the aetiology of RDS. Because the surfactant protein B (SP-B) is important for optimal surfactant function and because it is involved in the pathogenesis of pulmonary disease, we investigated the genetic variability of the SP-B gene in individuals with and without RDS. We identified a 2.5 kb BamHI polymorphism and studied its location, nature and frequency. We localized this polymorphism in the first half of intron 4 and found that it is derived by gain or loss in the number of copies of a motif that consists of two elements, a 20 bp conserved sequence and a variable number of CA dinucleotides. Variability in the number of motifs resulting from either deletion (in 55.3% of the cases with the variation) or insertion (44.7%) of motifs was observed in genomic DNAs from unrelated individuals. Analysis of 219 genomic DNAs from infants with (n = 82) and without (n = 137) RDS showed that this insertion/deletion appears with significantly higher frequency in the RDS population (29.3 as against 16.8%, P < 0.05).

Molecular analysis of a novel subtelomeric repeat with polymorphic chromosomal distribution

DNA from a 50-kb yeast artificial chromosome (YAC) containing one human telomere was characterized. Cloned sequences from the centromeric end of this YAC (designated yRM2001) localized to several human chromosomes by somatic hybrid panel mapping. The telomeric end of the YAC contained both (TTAGGG)n sequences and the previously characterized TelBam3.4 subterminal repeat element. A novel low-copy repeat element (designated HC1103) mapped 19 kb from the telomeric end of the YAC. This repeat was shown by fluorescence in situ hybridization to be present in several subtelomeric regions (3q, 12p, 15q, 19p, and 20p) and at an interstitial site (2q13-->q14) in all individuals studied, but to be polymorphically distributed at several other telomeres. The YAC vector-insert EcoRI cloning site was positioned 50 kb to 70 kb from chromosome termini in human genomic DNA using RecA-assisted restriction endonuclease (RARE) cleavage analysis. Our results suggest that the DNA segment cloned in yRM2001 contains a novel block of low-copy DNA consistently present at some human telomeres, but polymorphically distributed at others.

The Canon Group's effort: working toward a merged model

OBJECTIVE

To develop a representational schema for clinical data for use in exchanging data and applications, using a collaborative approach.

DESIGN

Representational models for clinical radiology were independently developed manually by several Canon Group members who had diverse application interests, using sample reports. These models were merged into one common model through an iterative process by means of workshops, meetings, and electronic mail.

RESULTS

A core merged model for radiologic findings present in a set of reports that subsumed the models that were developed independently.

CONCLUSIONS

The Canon Group's modeling effort focused on a collaborative approach to developing a representational schema for clinical concepts, using chest radiography reports as the initial experiment. This effort resulted in a core model that represents a consensus. Further efforts in modeling will extend the representational coverage and will also address issues such as scalability, automation, evaluation, and support of the collaborative effort.

Architectural requirements for a multipurpose natural language processor in the clinical environment

A considerable amount of research has been concerned with the development of natural language systems to automate the encoding of clinical information that occurs in textual form. The task is very complex, and not many language processors are used routinely within clinical information systems. Those systems that are operational, have been implemented in narrow domains for particular applications. For a system to be truly useful, it should be designed so that it could be widely used within the clinical environment. This paper examines architectural requirements we have identified as being necessary for portability and describes the architecture of the system we developed. Our system was designed so that it could be used in different domains to serve a variety of applications. It has been integrated with the clinical information system at Columbia-Presbyterian Medical Center where it routinely encodes clinical information from radiological reports of patients.

Physician advice to reduce chronic disease risk factors

The U.S. Preventive Services Task Force recommends that physicians advise their patients regarding smoking cessation, weight loss, and physical inactivity. Few studies, however, have assessed the extent to which persons with these risk factors receive advice from their physicians. Using data from the 1990-1991 Missouri Behavioral Risk Factor Surveillance System (BRFSS), a random digit-dialed telephone survey of adults, we identified Missouri residents with one or more of these modifiable risk factors. We examined whether these persons reported being advised by their physicians to modify their behavior(s) within the past year. Of the 2,791 respondents, 764 (26%) smoked, 1,720 (59%) were sedentary, and 686 (23%) were overweight. Five hundred and thirty-five smokers reported having a routine checkup within the past year, but only 224 (42%) reported being advised by their physicians to stop smoking. Of the 1,246 sedentary persons who had a routine checkup within the past year, 192 (15%) reported being told by their physicians to exercise more. Of the 521 overweight respondents who had a routine checkup within the past year, 225 (43%) reported being advised to lose weight. Physician advice for these risk factors was less frequently reported among men, blacks, younger persons, and persons from rural areas. Although most Missouri residents with these modifiable risk factors reported seeing their physicians within the past year, less than half reported that they received advice from their physicians to alter their risk behavior(s). Further efforts are necessary to increase the effectiveness of physician advice for at-risk patients about quitting smoking, losing weight, and increasing physical activity.

A general natural-language text processor for clinical radiology

OBJECTIVE

Development of a general natural-language processor that identifies clinical information in narrative reports and maps that information into a structured representation containing clinical terms.

DESIGN

The natural-language processor provides three phases of processing, all of which are driven by different knowledge sources. The first phase performs the parsing. It identifies the structure of the text through use of a grammar that defines semantic patterns and a target form. The second phase, regularization, standardizes the terms in the initial target structure via a compositional mapping of multi-word phrases. The third phase, encoding, maps the terms to a controlled vocabulary. Radiology is the test domain for the processor and the target structure is a formal model for representing clinical information in that domain.

MEASUREMENTS

The impression sections of 230 radiology reports were encoded by the processor. Results of an automated query of the resultant database for the occurrences of four diseases were compared with the analysis of a panel of three physicians to determine recall and precision.

RESULTS

Without training specific to the four diseases, recall and precision of the system (combined effect of the processor and query generator) were 70% and 87%. Training of the query component increased recall to 85% without changing precision.

A schema for representing medical language applied to clinical radiology

OBJECTIVE

Develop a representational schema for clinical concepts and apply it to the task of encoding radiology reports of the chest.

DESIGN

The schema was developed following a manual analysis of sample reports from the domain. The schema has two main components: the Medical Entities Dictionary (MED), which specifies the formal representation of the concepts in the domain and of their structures, and the natural-language processor, which specifies the linguistic expressions of the concepts. The schema was evaluated by applying it to a test set of 7,500 reports. Two-hundred reports from the test set were manually analyzed by a medical expert to determine the accuracy and success rate of the system.

RESULTS

82% of the 7,500 reports that contained relevant clinical information were successfully structured automatically. For the smaller set of 200 reports, 80% were structured successfully with an accuracy rate of 97%.

CONCLUSIONS

The schema is a formal representation for clinical concepts in radiology reports, and provides domain coverage that is particularly well-suited for natural-language processing of radiology for use in a decision support system.

Use of the total quality process in an infection control program: a surprising customer-needs assessment

The University of Michigan Hospitals began a quality management and improvement process in 1987 as the framework for all of its quality-related efforts and activities. The Infection Control Services department used total quality techniques to develop its mission statement, identify customers, identify customer requirements, and develop quality improvement objectives to meet the requirements. A service evaluation of customers resulted in specific improvement activities. An unexpected result of this evaluation was the difference noted between the Infection Control Services staff members' perception of customer requirements and these customers' actual needs. ICPs should use the continuous quality improvement tools and techniques to enhance their activities within their institutions, to better meet their customer needs, and to make sure that they are complementing their institution's mission.

A conceptual model for clinical radiology reports

The structural and informational content of clinical radiology reports was examined to develop a comprehensive representational schema of the concepts in the domain. The model involves several different conceptual levels, ranging from the high level description of the report to the lower level description of the clinical concepts contained in the reports and the specification of the terms used to express the concepts. The design of an adequate structured representation for the domain has important implications for the design of the electronic patient record, for the unification of different controlled vocabularies by enabling them to be mapped to one common representation, and for the facilitation of natural language processing of clinical reports so that coded data may be obtained.