The critical examination of clinical practice should be an integral part of patient care. It includes the development and implementation of guidelines, together with continuous evaluation of clinical process and outcomes to improve the quality of care provided. Clinical audit has not been successful in achieving this. The use of Integrated Care Pathways facilitates the introduction of guidelines and the continuous evaluation of clinical practice. Improvements are achieved by frequently revising the pathways to reflect current, local best practice. Integrated Care Pathways define the expected course of events in the care of a patient with a particular condition, within a set time-scale. A pathway is divided into time intervals during which specific goals and expected progress are defined, together with appropriate investigations and treatment. A pathway reflects the activities of a multidisciplinary team and can incorporate established guidelines and evidence-based medicine. It is usually unique to the institution in which it was developed. The pathway forms part of the clinical record of every patient. All variations from the pathway are documented, and the reasons for the variations analysed. Solutions are developed to address the causes of potentially avoidable variation, and the pathway is revised to incorporate these improvements. Integrated Care Pathways provide a powerful audit tool, as all aspects of the process and outcome of clinical practice can be constantly monitored. Variations from set standards are minimized, and improvements are rapidly incorporated into routine practice and subsequently re-evaluated.

This paper presents a mathematical model of Fraxinus pennsylvanica (green ash) shoots, which yields realistic computer graphics images of shoot geometry, and makes it possible to visualize shoot expansion using computer animation. The model is based on the mathematical formalism of Lindenmayer systems (L-systems). The paper emphasizes the process of model construction, with the aim of making the methodology applicable to other plant species. The central issue is the identification of field data necessary to build a comprehensive developmental model of plant architecture. Key words: Fraxinus pennsylvanica, plant architecture, shoot expansion, modeling, L-system, computer graphics.

Eleven rose (Rosa spp.) cultivars and advanced lines from the Morden rose breeding program were tested in controlled conditions for five components of partial resistance to two isolates of Diplocarpon rosae Wolf, the causal agent of black spot. The isolates were collected from diseased rose leaves in 1995 at Morden, Man., and in Ottawa, Ont. The components examined include incubation period (IP), leaf area with symptoms (LAS), number of lesions (NL), lesion length (LL), and sporulation capacity (SC). Differences in cultivar responses were found for all components. Of the 11 rose genotypes, no visible symptoms developed on line 91V8T201 even at 14 days after inoculation. This line was considered completely resistant to the two isolates tested. X6, T5, and `Prairie Joy' had a high level of partial resistance. N3 and `Cuthbert Grant' were intermediate. The remaining five cultivars or lines exhibited a low level of partial resistance. Linear correlations among LAS, IP, NL, and LL were high. Correlations between SC with IP, NL, or LL were nonsignificant. Although each of the five components can be used in selection for partial resistance, any one component may not represent the complete potential for a particular rose genotype. Results of this study suggest that LAS and SC are preferable as criteria for resistance to minimize the labor requirement when evaluating large numbers of rose genotypes in the greenhouse.

Summary. Shoot preformation was investigated in buds of four clones of Fraxinus pennsylvanica var. subintegerrima (Vahl) Fern. at two sites in Manitoba in the second (1988) and third (1989) growing seasons after grafting. More preformed primordia were produced in terminal buds in 1989 compared to 1988 at each site. Both terminal and lateral buds at Morden contained significantly more primordia than those at Winnipeg. The numbers of preformed primordia were significantly different among clones. Clone 3 produced the most and clone 1 the fewest primordia in terminal buds. Despite quantitative variation, the pattern was similar among clones for terminal buds at each site and in each year. A similar pattern was evident for lateral buds at the Winnipeg site in 1989 but at Morden, clones 4 and 1 had the largest number of preformed primordia. Data from 1989 revealed that numbers of primordia were correlated with bud dimensions, parent shoot length, diameter and number of leaves, and location of the bud on the parent. Shoot dry weight was also related to these variables and revealed a non-linear increase in dry weight with shoot length. Multiple regression, with parent shoot length and location of buds along the parent axis as independent variables provided a reliable indicator of preformation in the crown. Although there is a genotypic component to preformation, variation between sites, years and crown locations suggests plasticity in bud development.

No abstract