U. Plaul

Outcome nach minimal-invasiver Chirurgie Qualitative Analyse und Bewertung der klinischen Relevanz von Studienendpunkten durch Patient und Arzt

Abstract.Introduction: Mechanistic study endpoints, evaluated exclusively by the physician, are mostly used in clinical studies evaluating new treatment modalities (e.g. laparoscopic cholecystectomy). Those endpoints often lack clinical relevance. The patients opinion concerning the importance of a study endpoint is particularly important in the evaluation of minimally invasive procedures, which place special emphasis on patient comfort. Methods: In a first step it was evaluated by meta-analysis, which clinical endpoints have been used for comparison of laparoscopic and conventional cholecystectomy. Furthermore, using a qualitative analysis it was investigated how important the individual study endpoints are for patients and physicians. Ten patients and five surgeons were questioned in a structured interview. Results: Of all outcome variables used world-wide, approximately one third were hermeneutic study endpoints, depending on the quality of the study, but often the method of evaluation was insufficient. Only three of 215 endpoints ( < 2 %) were quality of life scores, an integrated concept of outcome was missing completely. The qualitative analysis confirms the claimed difference between isolated and integrated evaluation of treatment goals. The importance of postoperative death is underestimated by patients and physicians; postoperative pain is overestimated. Patients ranked the outcome variable “restoration of full physical fitness“ as the most important study endpoint after avoidance of complications and death. It is underestimated in isolated evaluation and has not been used in the world literature at all. Conclusion: The analysis of clinical relevance of study endpoints should be the first and not the last step of studies to evaluate surgical technology. It cannot be based purely on intuition; it must make use of scientifically accepted techniques (e.g. qualitative analysis).Zusammenfassung.Einleitung: In klinischen Studien zur Evaluation neuer Behandlungsverfahren (wie z. B. bei der Einführung der laparoskopischen Cholecystektomie) werden zumeist mechanistische, allein vom Arzt evaluierte Studienendpunkte gewählt. Häufig fehlt solchen Endpunkten die klinische Relevanz. Gerade bei der Beurteilung von minimal-invasiven Behandlungsmethoden, bei denen der Patientenkomfort im Vordergrund steht, ist die Beurteilung eines Studienendpunkts durch Patienten besonders wichtig. Methoden: Es wurde zunächst im Rahmen einer Metaanalyse evaluiert, welche klinischen Studienendpunkte beim Vergleich von laparoskopischer und konventioneller Cholecystektomie bisher verwendet wurden. Anschließend wurde mit einer qualitativen Analyse untersucht, welche Bedeutung den einzelnen Studienendpunkten von Patienten und Ärzten beigemessen wird. Hierzu wurden 10 Patienten und 5 Chirurgen in einem strukturierten Interview befragt. Ergebnisse: In Abhängigkeit von der Studienqualität wurden weltweit zu etwa einem Drittel hermeneutische Gesundheitsziele (Endpunkte) eingesetzt, aber vielfach mit ungenügender Methode. Lebensqualitätsscores wurden als 3 von 215 Endpunkten verwendet ( < 2 %), ein integratives Konzept von Outcome fehlte vollständig. Die qualitative Analyse weist den behaupteten Unterschied zwischen isolierter und integrativer Bewertung von Gesundheitszielen nach. Der postoperative Tod wurde von Patient und Arzt unterschätzt, der postoperative Schmerz überschätzt. Das von Patienten nach Vermeidung von Komplikationen und Tod am höchsten bewertete Gesundheitsziel „Wiederherstellung der vollen physischen Belastbarkeit“ wurde bei isolierter Wertung unterschätzt und tauchte in Studien der Weltliteratur überhaupt nicht auf. Schlussfolgerung: Eine Werteanalyse der klinischen Relevanz von Studienendpunkten muss an den Beginn, nicht ans Ende von Studien zur chirurgischen Technologiebewertung gestellt werden. Diese darf nicht allein aus Intuition kommen, sie muss sich heute wissenschaftlich anerkannter Methoden (z. B. qualitativer Analyse) bedienen.

Dependence of Positive Effects of Granulocyte Colony-stimulating Factor on the Antibiotic Regimen: Evaluation in Rats with Polymicrobial Peritonitis

We tested the hypothesis that the ability of granulocyte colony-stimulating factor (G-CSF) to prevent death from fecal peritonitis is influenced by the composition of the antibiotic regimen with which it is administered. We used a rodent model of polymicrobial peritoneal contamination and infection and the concept of clinical modeling randomized trials (CMRTs), which includes the conditions of randomized, clinical trials and complex clinical interventions (e.g., anesthesia, volume substitution, antibiotics, surgery, postoperative analgesia). With the peritonitis model we obtained a mortality dose-response curve that was sensitive to antibiotic prophylaxis. G-CSF was most efficacious when it was administered both prophylactically and after the onset of peritonitis. Cefuroxime/metronidazole, ofloxacin/metronidazole, and amoxicillin/clavulanate improved survival in combination with G-CSF best, whereas cefotaxime or ceftriaxone with and without metronidazole did not. G-CSF administration was associated with improved polymorphonuclear neutrophil phagocytosis and enhanced bacterial clearance. Pro-inflammatory cytokine release (tumor necrosis factor-a, interleukin-6, macrophage inflammatory protein-2) was decreased in plasma and in the peritoneal fluid. Their expression was lowered in various organs on the protein and mRNA level. The results were used to design a clinical trial to test the ability of G-CSF to prevent serious infections in patients with colorectal cancer surgery. In this trial G-CSF application and antibiotic prophylaxis were performed with the most effective scheduling and combinations (cefuroxime/metronidazole and ofloxacin/metronidazole) as defined here.

Granulocyte-colony stimulating factor in the prevention of postoperative infectious complications and sub-optimal recovery from operation in patients with colorectal cancer and increased preoperative risk (ASA 3 and 4)¶ Protocol of a controlled clinical trial developed by consensus of an international study group¶ Part three: individual patient, complication algorithm and quality management¶

Abstract:General design: Presentation of a new type of a study protocol for evaluation of the effectiveness of an immune modifier (rhG-CSF, filgrastim): prevention of postoperative infectious complications and of sub-optimal recovery from operation in patients with colorectal cancer and increased preoperative risk (ASA 3 and 4). A randomised, placebo controlled, double-blinded, single-centre study is performed at an University Hospital (n = 40 patients for each group). This part presents the course of the individual patient and a complication algorithm for the management of anastomotic leakage and quality management.¶Objective: In part three of the protocol, the three major sections include:¶ - The course of the individual patient using a comprehensive graphic display, including the perioperative period, hospital stay and post discharge outcome.¶ - A center based clinical practice guideline for the management of the most important postoperative complication - anastomotic leakage - including evidence based support for each step of the algorithm.¶ - Data management, ethics and organisational structure.¶Conclusions: Future studies with immune modifiers will also fail if not better structured (reduction of variance) to achieve uniform patient management in a complex clinical scenario. This new type of a single-centre trial aims to reduce the gap between animal experiments and clinical trials or - if it fails - at least demonstrates new ways for explaining the failures.¶

Effects of G-CSF and antibiotic prophylaxis in a 2×2 factorial design on outcome in septic rats

AbstractObjective:In a recently completed randomised clinical trial in patients with colorectal cancer resections the combination of the granulocyte-colony stimulating factor (G-CSF) + cefuroxime/ metronidazole (cef/met) was superior to ofloxacin/metronidazole (ofl/met). These combinations were used to confirm the clinical data and to validate the concept of clinic modelling randomised trials (CMRTs) in a rat model of intra-abdominal sepsis. Subjects:80 male Wistar rats were randomised in a 2×2 factorial study design. Treatment:All animals (n = 20/group) received anaesthesia, antihistamines, antibiotic prophylaxis, peritoneal contamination and infection. Groups were: 1) G-CSF+cef/met; 2) placebo+cef/ met; 3) G-CSF+ofl/met; 4) placebo+ ofl/met. G-CSF (20 g/kg) prophylaxis was applied three times. Methods:Survival at 120 h was analysed with the Kaplan Meier method. Results:Survival rate was best in the G-CSF+cef/met group with 75% and was significantly improved compared to the cef/met placebo group, in which only 42% survived (P < 0.05). Survival rate between both G-CSF groups was similar being 75% in the cef/met and 72% in the ofl/met group. P = 0.10). Ozone increased TNF-α and MIP-2 after infection: 127±23 pg/ml and 94±19 pg/ml (control group: 398 pg/ml and 369 pg/ml; P < 0.002 and P < 0.01). IL-6 levels were similar in both groups. Conclusions:The results of this CMRT confirmed the result of our clinical G-CSF trial in that G-CSF prophylaxis was most efficacious in combination with cef/met to improve the outcome.

The genetic background of hypertensive, septic rats determines outcome improvement with antibiotic and G-CSF prophylaxis

Hypertension is proposed as a risk factor among others (high age, diabetes mellitus, and pre- and intraoperative bleeding) for adverse outcomes, such as severe infections, leading to sepsis and to multiple organ failure as the most deleterious complication. Hypertension was modeled with spontaneous hypertensive rats (SHR) and Dahl salt-sensitive (DS) rats and the infective complication by polymicrobial, peritoneal contamination, and infection (PCI). The concept of clinic modeling randomized trials was used to simulate clinical complexity, including a relevant antibiotic prophylaxis in combination with granulocyte-colony stimulating factor (G-CSF) and clinical trial conditions. Outcome parameters were: survival, systemic cytokines (protein), and organ-specific cytokine levels (mRNA). With low complexity (no prophylaxis), 28% of the animals in the Wistar and 50% in the SHR group survived (P = 0.17). Tumor necrosis factor-&agr; levels were lower in the liver of SHR vs. Wistar rats with PCI (P < 0.01). The anti-inflammatory cytokine interleukin (IL)-10 was expressed on a higher level in SHR with PCI compared with Wistar rats (P < 0.01). With increased complexity (antibiotic and G-CSF prophylaxis) the survival rate was increased from 50% in Wistar rats to 89% in SHR (P < 0.01) and the mRNA expression of IL-6 was decreased in the kidney of SHR (P < 0.05). Survival rate was 44% in the DS rats vs. 67% of the Wistar rats (P = 0.18). The mRNA expression of tumor necrosis factor-&agr; and IL-10 was reduced (P < 0.01) by pretreatment in the liver of DS rats with PCI. The hypertensive, genetically distinct SHR and DS rats express different patterns of pro- and anti-inflammatory cytokine levels after PCI. G-CSF and antibiotic prophylaxis increases only in SHR survival and decreases IL-6 mRNA expression in the kidney significantly.

Computergestützte Visualisierung von Daten der Lebensqualität zur Therapieoptimierung am Beispiel von Patienten mit kolorektalem Karzinom

Introduction: A variety of questionnaires (EORTC-QLQ30, SF-36 a. o.) are available to assess quality of life (QL). A lot of data are collected with these questionnaires but the interpretation of these data is very complex. Methods: We developed a method to visualize these complex data: the QL profile. In this profile different questions composing a particular QL dimension were aggregated and linearly transformed to a scale from 1 to 100. The resulting score values were graphically arranged to a comprehensive QL profile [1]. This procedure was repeated for each QL measurement (e.g. 2, 6, 9 etc. month after surgery). Results: We showed that these profiles are understandable for physicians not educated in QL measurement. The interpretation of the profiles sometimes leads to a different treatment of patients and shows that it is possible to control the quality of treatment any point of time. Conclusion: With our QL profiles it is possible to visualize complex data from QL-questionnaires in an intuitive comprehensible way. Therefore, QL questionnaires can become a routine tool for physicians to follow-up their patients.

G-CSF verbessert das postoperative Outcome von Hochrisikopatienten nach Operation aufgrund eines kolorektalen Karzinoms

Introduction: Recombinant human granulocyte colony stimulating factor (G-CSF, filgrastim) is a cytokine which is commonly used to prevent neutropenia and its complications in tumor patients. Recently in a systematic review was shown that G-CSF did not alter the survival rate and number of remissions in patients with malignant lymphoma, but G-CSF showed a positive influence on accompanying diseases such as infections [1]. The aim of this trial was to test the effectiveness of a G-CSF prophylaxis in high risk patients with regard to improvement of the postoperative outcome [2]. Methods: To analyze this question we performed a prospective, randomized, multicenter trial with 80 high risk patients (ASA 3 and 4) and colorectal cancer surgery [2]. The patients were randomized in a 2 × 2 factorial design to: 1) G-CSF plus cefuroxime/metronidazole (Cef/Met), 2) G-CSF plus ofloxacin/ metronidazole (Ofl/Met), 3) placebo plus Cef/Met, and 4) placebo plus Ofl/Met. G-CSF (300 µg) was injected subcutaneously 12 h before operation, and 12 and 36 h after surgery. Quality of life (QoL) [4] was assessed before operation, at discharge, 2 and 6 months postoperatively with the EORTC-QLQ-C30 and the colorectal cancer module C38 [4]. Results: In the study time of 2 years 162 patients with ASA 3 and 4 had an left-sided colorectal cancer resection in the three study centers. From this 162 patients 80 fulfilled the inclusion criteria. Baseline demographic and physiological data (age, gender, ASA, nutritional status, concomitant disease, type of operation) were not different between the groups. At hospital discharge global QoL, family life, negative affect, physical functioning, cognitive functioning was best and pain lowest in the G-CSF plus cef/met group (optimum 100 QoL points) (¤ Table 1). Furthermore patients in the cef/met group had the shortest length of hospital stay and the fewest numbers of complications. Conclusions: G-CSF prophylaxis improves the postoperative quality of life in high risk patients with colorectal cancer resections. This G-CSF effect is dependent on the antibiotic used. Clinical complexity interactions are decisive for the effectivity of this prophylaxis.

Was ist für Patienten mit einem kolorektalen Karzinom wirklich wichtig? Prospektive Analyse von 80 Patienten mittels eines strukturierten Interviews

Postoperative outcome of patients with colorectal cancer surgery is usually only assessed by mechanistic outcome variables (mortality and morbidity). In contrast, we developed an integrated outcome concept. This concepts includes doctor assessed physical functioning and patient self-reported quality of life and a value judgement of both [1]. For value judgement we tested a structured interview, to obtain further information of clinical relevance beyond that of quality of life questionnaires. Methods: A 6 month follow up was performed in the course of a prospective, double blinded, randomised trial with G-CSF prophylaxis for improvement of postoperative outcome of patients after colorectal cancer surgery [2]. We assessed the objective health status, patients expressed quality of life with questionnaires and the personal experience in an interview lasting about 20 minutes [3]. The interview consisted of an introduction, the possibility to talk about all cancer associated experiences and a structured interview with six questions: 1) How is your general health condition? 2) Are you recovered to the level before operation? 3) What was the worst/ 4) the best/ 5) the most important events for you during the last six months? 6) Where there any major changes in your life circumstances? The interview was tape-recorded for further analysis. Results: From the 80 patients randomised in the trial 63 interviews were obtained. One patient was not operated, 6 died and 10 were either nor able or refused it. Six months after operation 75% of the patients claimed to be in a good or acceptable shape. The worst experience was for 16% the diagnosis cancer, for 14% the operation and for 13% the radio-chemotherapy.

Klinische Relevanz von Studienendpunkten randomisierter klinischer Studien: Metaanalyse und qualitative Analyse am Beispiel des Vergleichs von laparoskopischer vs. konventioneller Cholezystektomie

Introduction: To choose one or more appropriate study endpoints is one of the most important steps in planning randomized clinical trials. The choice is made by physicians and statisticians for various reasons, one being feasibility. The evaluation by the patients themselves is rarely and only indirectly taken into consideration. Therefore, the clinical relevance of many study endpoints must be considered as questionable. Using the example of the comparison of laparoscopic and conventional cholecystectomy, the degree of importance attached to the different goals of medical treatment by patients and physicians was systematically evaluated. Methods: In total, 18 randomized trials comparing laparoscopic and conventional cholecystectomy were identified in a meta-analysis. The study endpoints of these trials were quantitatively ascertained and used in a subsequent qualitative analysis (Dey 1993). On the day prior to elective cholecystectomy ten patients and five surgeons performing the operations were asked in a standardized interview about the relevance which they attached to these endpoints: initially as single variables using a Likert scale, and then in a competitive sequence (rank list with cards). Results: The classical mechanistic outcomes such as mortality, complication rate and length of in-hospital stay were almost exclusively identified in the meta-analysis. Hermeneutic endpoints (reported by the patient from his subjective experience) were rarely used except for postoperative pain, the latter often with invalid methods. Quality-of-life scores were only used as two of a total of 72 endpoints (< 3%), an integrated concept of outcome (patient and physician) was missing completely. Rating the different outcomes (qualitative analysis) surgeons as well as patients considered death and intra- and postoperative complications as most important. Return to full physical fitness was the most important hermeneutic endpoint for the patients, while this endpoint was much less important for the surgeons. Most patients considered postoperative pain as much less important than physicians. Length of hospital stay was often given the lowest rank. Conclusion: An integrated outcome concept is recommended which includes the rating of study endpoints by patients and physicians. Analysis of the clinical relevance of outcome variables needs to be one of the first and not the last steps in surgical technology assessment. It should not originate exclusively from the intuition of the doctor, but needs to be done with new scientifically accepted methods (e.g. qualitative analysis).