تجزیه و تحلیل به حداقل رساندن هزینه ترکیبی از تکنیک های نمونه گیری در برونکوسکوپی ضایعاتendobronchial

Background The choice of sampling techniques in bronchoscopy with sampling from a visible lesion will depend on the expected diagnostic yields and the costs of the sampling techniques. Aims The aim of this study was to determine the most economical combination of sampling techniques when approaching endobronchial visible lesions. Methods A cost minimization analysis was performed. All bronchoscopies from 2003 and 2004 at Haukeland university hospital, Bergen, Norway, were reviewed retrospectively for diagnostic yields. 162 patients with endobronchial disease were included. Potential sampling techniques used were biopsy, brushing, endobronchial needle aspiration (EBNA) and washings. Costs were estimated based on registration of equipment costs and personnel costs. Sensitivity analyses were performed to determine threshold values. Results The combination of biopsy, brushing and EBNA was the most economical strategy with an average cost of Euro 893 (95% CI: 657, 1336). The cost of brushing had to be below Euro 83 and it had to increase the diagnostic yield more than 2.2%, for biopsy and brushing to be more economical than biopsy alone. The combination of biopsy, brushing and EBNA was more economical than biopsy and brushing when the cost of EBNA was below Euro 205 and the increase in diagnostic yield was above 5.2%. Conclusion In the current study setting, biopsy, brushing and EBNA was the most economical combination of sampling techniques for endobronchial visible lesions.

Bronchoscopy is the main diagnostic method for endobronchial visible lesions.1, 2 and 3 A biopsy allows for a histological sample, which inherently has more information than a cytological sample. However, in many instances, a cytological sample can be sufficient for the establishment of a malignant diagnosis in the airways, and most physicians will prefer to take cytological samples in addition to biopsies when encountering an endobronchial lesion. The choices for cytological sampling are brushing, endobronchial needle aspiration (EBNA), and washings. The optimal combination of all sampling techniques is still debated, due to differences in yield and cost. The previous recommended combinations have been biopsy, brushing and washing,4, 5, 6 and 7 biopsy and brushing,8, 9 and 10 biopsy, brushing or washing,11 and 12 biopsy, brushing and EBNA,13, 14 and 15 or EBNA alone with rapid on-site cytological evaluation (ROSE) combined with supplemental techniques if negative.16 So far, the diagnostic yield has been the most important criterion for the recommendations. However, the costs of the different sampling techniques vary greatly, and the overall cost of the procedure will in a large part be dependent on how each center organizes its diagnostic procedures. The main weakness of the current recommendations is the lack of data on true costs. Although some studies of bronchial washings have included costs,11 and 12 the most economical combination of all sampling techniques has yet not been settled. The aim of this study was to determine the most economical combination of sampling techniques in bronchoscopy of endobronchial visible lesions, using a cost minimization analysis.

In this study, the combination of biopsy, brushing and EBNA in one procedure was the most economical strategy for diagnosing endobronchial lesions, with an average cost of Euro 893 compared to Euro 970 for biopsy and brushing and Euro 999 for biopsy alone. The main strength of this study was the application of an analysis that involved costs in the comparison of different diagnostic yields. The results and the threshold values should be comparable to other institutions. There are some methodological issues to consider. Firstly, all sampling techniques were applied simultaneously in only 43 of 162 cases. The choice of combinations of sampling techniques was left to the investigator. If it was difficult to reach a lesion, fewer sampling techniques might have been applied, thus overestimating the diagnostic value of multiple sampling techniques. On the other hand, some investigators would perhaps use more sampling techniques in hard-to-reach-lesions. Secondly, the sample size of this study was too small to allow adjustment for the size of the lesion or the location. Thirdly, in prospective studies it would be possible to collect data about all costs for each patient. In a retrospective study such as this, the costs must be based on a model with average costs for each procedure, which is less precise. In this study the model for costs was based on registration of the time consumption for different health workers in the bronchoscopy lab and in the department of pathology. The time consumption is not constant and the variation was only included in the assumption that the costs were in a range of a 30% deviation. The time consumption in the cytopathology lab was based on expert opinion, prone to information bias. These assumptions and uncertainties of costs must be considered, especially when comparing to costs reported in other studies. EBNA is the most costly sampling technique. However, prolonging the stay at the day ward and having a new procedure performed is much more costly. Thus, adding EBNA to all procedures seems warranted from the modeling in this study, since this prevented a large enough number of repeated procedures to be less costly overall. Whether this would be true for other centers will depend upon the local diagnostic yield for each procedure, and the organization of the diagnostic work-flow until conclusion. The effectiveness of different combinations of sampling techniques has been an issue in previous studies.4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 and 16 In the current study, the combination of biopsy and brushing was more economical than biopsy alone given that the diagnostic yield for the combination was at least 78.0% compared to the diagnostic yield of biopsy alone of 75.8%. Previous studies have reported that brushing increased sensitivity for cancer above 2.2% compared to biopsy alone.5, 8, 11, 14, 15, 22 and 23 Compared to the diagnostic yield of biopsy and brushing of 78.8%, the combination of biopsy, brushing and EBNA was most economical in the current study if the diagnostic yield was above 84.0%. Few previous studies have included EBNA in the analyses. Three studies reported increased sensitivity for EBNA above 5.2% compared to biopsy and brushing.13, 14 and 15 Given the diagnostic yields in the current study, the cost of brushing had to be below Euro 83, if biopsy and brushing should be more economical than biopsy alone. One previous study had a price of brushing below this threshold value,12 while in the other study the price of brushing was above this threshold value.11 The cost of EBNA had to be below Euro 205, for the combination of biopsy, brushing and EBNA to be more economical than biopsy and brushing. Only one study included EBNA; with a price below the threshold value.12 Although the current study recommends biopsy, brushing and EBNA in one procedure, the results must be interpreted cautiously. Some institutions investigate the patients as outpatients to a much larger degree than ours. If the cost of a repeated bronchoscopy including the cost of the delay is below Euro 1786, biopsy and brushing without EBNA might be a more economical choice for an outpatient investigation. In countries where the cost for the waiting time is low, biopsy alone might be most economical. Aspiration of fluid from the entire procedure for washings did not increase the diagnostic yield regardless, and has been omitted in our hospital for the investigations of visible lesions. For the clinical decision maker the optimal combination of sampling techniques will be based on the willingness to pay for increased diagnostic yield. A great increase may warrant the extra costs of more sampling techniques as that would save the time and cost for a new procedure. If a center knows its own diagnostic yields for the different sampling techniques, a sensitivity analysis can reveal the threshold values for when a given technique becomes economical. It is important to emphasize that the current analysis is economical in nature, and does not take into account the psychological burden of time to diagnosis. We do not advocate to prolong the investigative time to decrease costs. On the contrary, the analyses show that even though the increased diagnostic yield was small with added sampling techniques employed, it saved money due to the high cost of diagnostic delay. For the diagnosis of lung cancer, each diagnostic center will have slightly different results. The costs will vary with different practices and models of financial support. Thus, we recommend all centers, both excellent and regular, to calculate both their diagnostic yields, and the costs of their diagnostic work-ups, and finally to publish these results.