Options on fibroid morcellation: a literature review.

Given the scarcity of direct morcellation injuries, no clinical trials are available and all suggestions to prevent them may be considered as good practice points. With regards to preventing direct morcellation injuries, the options are listed in Table  7 ; regarding the development of parasitic fibroids, the options are mentioned in Table  8 . The options to prevent upstaging of uterine sarcomas by power morcellation will be presented under the final options in Table  10 . Table 7 Options to prevent direct morcellation injuries Options Grade For safe entry, enlarge the skin and fascia incision to the diameter of the morcellator to reduce the abdominal wall resistance Good practice point Make sure that the morcellator’s blade remain locked inside the protecting tube during the morcellator insertion into the abdomen Good practice point Keep the tip of the morcellator shaft in midline of the lower abdomen while introducing the device into the abdominal cavity and during morcellation Good practice point Morcellate only under continuous vision by applying the lateral pealing technique. Prevent penetrating the mass and losing the tip out of sight Good practice point Morcellation close to the intestine or to blood vessels increase risk of injury to these structures Good practice point Table 8 Statements and options on preventing parasitic fibroids after morcellation Statements and options Grade The small risk of parasitic fibroid with laparoscopic morcellation (<1 %) should be discussed with the patient and balanced against alternative treatment options Good practice point Avoid spread of cells and tissue fragments in the abdominal cavity by stabilising the specimen and prevent fast rotation Good practice point When morcellation is used, efforts should be made to prevent tissue loss during morcellation and to remove all tissue fragments after morcellation: Place the patient in reverse Trendelenburg position after morcellation and irrigate the abdomen and pelvis extensively After irrigation of the peritoneal cavity the abdomen and pelvis should be inspected to identify any remaining tissue fragments Good practice point The potential increased risk of parasitic fibroids after sex steroid exposure (endogenous/exogenous) after laparoscopic morcellation should be considered before hormonal replacement therapy is prescribed D Options to prevent direct morcellation injuries Statements and options on preventing parasitic fibroids after morcellation When morcellation is used, efforts should be made to prevent tissue loss during morcellation and to remove all tissue fragments after morcellation: Place the patient in reverse Trendelenburg position after morcellation and irrigate the abdomen and pelvis extensively After irrigation of the peritoneal cavity the abdomen and pelvis should be inspected to identify any remaining tissue fragments

A complication can be defined as an unintended and undesirable event following clinical management resulting in its adjustment or irreversible injury to the patient. Known complications are direct morcellation injuries where the activated morcellator injures intestines or blood vessels. Secondary to morcellation of fibroids, chips can implant on the peritoneum causing parasitic fibroids which may need further surgery. If a presumed fibroid appears to be a sarcoma (or other malignancy), any method of morcellation disrupts the integrity of the tumour, possibly upstaging the disease and affecting survival. In case of power morcellation, the centripetal forces of the cylindrical knife may add to the phenomenon of ‘seeding’ of tumour cells on the peritoneum. Finally, the fragmental state of the specimen due to morcellation may impair proper selection for histologic evaluation of that part of the tumour that is suspicious of malignancy. This might cause treatment delay. In this section, these complications will be discussed separately. In the field of gynaecology, the majority of symptomatic masses are represented by uterine fibroids. During operative laparoscopy, power morcellation is indispensable to remove large tissue fragments while the laparoscopy offers quicker recovery, less post-operative pain, fewer wound complications and less post-operative morbidity than open procedures. In addition, power morcellators shorten the time of surgery significantly. An ‘electrical cutting device for laparoscopic removal of tissue from the abdominal cavity’ was introduced by Steiner in 1993 [ 71 ]. Similar instruments according to the Steiner principle have been commercialised since. Despite the well-established advantages of power morcellation during laparoscopy, the use of power morcellators is not completely without concern. Milad [ 72 ] reviewed the FDA’s adverse event database “MAUDE” between 1992 and 2013, where injuries to the small/large bowel (31), large blood vessels (27), the kidney (3), ureter (3), bladder (1) and diaphragm (1) have been reported using power morcellation. In six cases, the accidents were fatal. This underlines that direct morcellation injuries are serious and, though underreported, extremely rare. It is recommended to maintain adequate distension and use morcellators with a nozzle to promote lateral pealing preventing the morcellator from coring the tissue and thereby losing the morcellator’s tip out of sight [ 73 ]. Traditionally, parasitic fibroids were thought to be pedunculated subserosal fibroids that were accidentally separated from the uterus and had become attached to another organ in the pelvis for their blood supply (3). The increasing number of case reports of parasitic fibroids after the use of laparoscopic morcellation has contributed to the development of an iatrogenic theory. It is thought that seeding of retained small tissue fragments after morcellation can lead to the development of parasitic fibroids in the peritoneal cavity [ 74 ]. Most patients presented with symptoms such as abdominal or pelvic pain, dyspareunia, abdominal distension, urinary frequency and constipation. One of five patients is asymptomatic and the presence of a pelvic mass unexpectedly diagnosed during routine examination or another surgical procedure. Additional published cases indicated that although rare, parasitic fibroids can occur long after laparoscopic morcellation and are often asymptomatic or present with abdominal or pelvic pain. Little is known about the incidence and risk factors of this phenomenon. To answer questions about the incidence, risk factors and other characteristics of parasitic fibroids, one of the authors (HP) conducted a systematic literature search the MEDLINE and Embase databases. The report of this systematic review will be submitted elsewhere. The overall incidence of parasitic fibroids after laparoscopic surgery with the use of morcellation was reported to be between 0.12 and 0.9 % [ 75 – 77 ]. The reported incidence of parasitic myomas after laparoscopic myomectomy was 0.2–1.2 % [ 76 , 78 , 79 ]. With regards to the risk factors , gonadal steroids hormones are known to influence the growth of uterine fibroids. It is hypothesised that prolonged exposure to steroid hormones (e.g. hormonal replacement therapy) during postmenopause could be a risk factor for the development of parasitic fibroids. In case of parasitic fibroids, there is often more than one. Statements on parasitic fibroids are listed in Table  4 . Table 4 Statements on parasitic fibroids by previous morcellation Statements Grade The overall incidence of parasitic fibroids after laparoscopic surgery with the use of morcellation is reported to be between 0.12 and 0.9 % D The reported incidence of parasitic fibroids after laparoscopic myomectomy is 0.2–1.2 % D Premenopausal status and hormonal replacement treatment after primary surgery may be considered as risk factors for the development of parasitic fibroids, however not specific D Statements on parasitic fibroids by previous morcellation Although reducing a uterine sarcoma with electromechanical power morcellation within the abdominal cavity is contrary to oncologic surgical principles, the question remains if sarcoma cells are more inclined to implant in the peritoneum after power morcellation then after ‘en bloc’ removal of the uterine specimen with sarcoma embedded and if survival is affected given the bad prognosis that the sarcoma already has. These two questions have been addressed in literature. Two case reports support the concept of upstaging by power morcellation [ 80 , 81 ]. In the reported patients, reintervention showed seeding of sarcomatous tissue, which was not visible during initial surgery. These findings have been confirmed in larger studies where the percentage upstaging ranged from 15 to 64 % [ 18 , 28 , 82 , 83 ]. Not only electromechanical power morcellation is associated with the risk of upstaging but also other ‘manipulations’ of the sarcomatous tumour, such as myomectomy by laparotomy, subtotal hysterectomy and hysteroscopic resection of submucous fibroids, may affect survival suggesting upstaging [ 84 ]. The question as to whether seeding affects survival is also addressed by Seidman et al. [ 18 ]. In four of seven patients with LMS who were extracted by power morcellation, seeding was visible during reintervention. Three patients died from the disease and one was alive with metastatic disease. Three patients with morcellated sarcoma without signs of seeding were alive without metastases. Two studies by Park et al. [ 85 , 86 ] compared the survival of patients with uterine sarcoma with ( n  = 48) and without morcellation ( n  = 58) during surgery and demonstrated a significant difference of survival in favour of the non-morcellated group (Table  5 ). Although confounding factors cannot be excluded in these retrospective trials, an effect in favour of the morcellation group would have been expected as clinical suspicion of sarcoma based on size and imaging texture could have resulted in a worse prognosis in the laparotomy group. Table 5 Studies that compared patients operated for uterine sarcoma with and without morcellation. With permission from Nederlands Tijdschrift voor Obstetrie en Gynaecologie [ 27 ] Park 2011 (LGESS) No. Age FU Recurrence 5 years DFS ORmv Morcellation − 27 45.3 64 3/27 84 % Morcellation + 23 43.6 66 8/23 55 % 4.03 (1–15) Park 2011 (LMS) n Age FU Recurrence 5 years DFS OR Morcellation − 31 47.9 52 7/31 65 % Morcellation + 25 46.4 27 13/25 40 % 3.11 (1–9) LGESS low grade endometrial stromal sarcoma, LMS leiomyosarcoma, FU follow-up, DFS disease free survival, ORmv odds rate mortality risk after morcellation in a multivariate analysis Studies that compared patients operated for uterine sarcoma with and without morcellation. With permission from Nederlands Tijdschrift voor Obstetrie en Gynaecologie [ 27 ] LGESS low grade endometrial stromal sarcoma, LMS leiomyosarcoma, FU follow-up, DFS disease free survival, ORmv odds rate mortality risk after morcellation in a multivariate analysis Recently, a literature review was published by Pritts et al. [ 87 ] where she critically appraised the studies that reported on upstaging of sarcoma by morcellation which she evaluated as ‘rather poor’. Although this weakens the alarming statements on upstaging, it does not prove the contrary (the innocence of open morcellation of sarcoma) and underlines the need for further studies. Statements on upstaging of uterine sarcoma by morcellation are listed in Table  6 . Table 6 Statements on the complication of morcellation ‘seeding’ (upstaging uterine sarcoma) Statements Grade The quality of research regarding upstaging of uterine sarcoma by open morcellation is rather poor D Electromechanical power morcellation of an unsuspected uterine sarcoma may cause intraperitoneal dissemination (‘seeding’) C Intraperitoneal dissemination (‘seeding’) may be associated with lower survival rates C ‘En bloc’ resection of a uterine sarcoma may be associated with better survival than other tissue retrieval methods going with tumour injury D Statements on the complication of morcellation ‘seeding’ (upstaging uterine sarcoma) Morcellated specimens are poorly amenable to pathologic examination because the morcellation abolishes many of the anatomic features that allow meaningful gross description, including the notions of orientation, dimension, adjacency, border and margin [ 88 ]. This has been described in other uterine malignancies and could lead to delayed diagnosis or suboptimal staging, causing treatment delay [ 89 , 90 ].

The overall risk of not previously presumed sarcomatous change in the uterus from all papers was 0.14 % (1 in 700). However, there were large differences between papers with figures varying from 0.49 % (1 in 204) [ 19 ] to 0.056 % (1 in 1,788) [ 16 ]. On average, papers that looked at myomectomy specimens gave a lower risk of sarcomatous change of 0.08 % (1 in 1,306) compared to those that looked at hysterectomy specimens where the overall pooled risk was 0.15 % (1 in 650). The risk appears to be age related with one study demonstrating a lower risk in patients under 45 years of age (Ind et al. 2014, personal communication).

The—on some issues—systematic literature search on which this review is based revealed some serious gaps in available knowledge. Due to the low prevalence of uterine sarcomas, most data on prevalence and risk factors are derived from case histories or retrospective trials with low numbers. Prospective data collection of patients after fibroid surgery on age, imaging and laboratory results and subsequent histology is needed to answer questions on individual risk estimation in order to enable the patient to give a well-considered consent prior to minimal invasive surgery. The development of safer morcellation techniques in the abdominal cavity by technical innovation, including in-bag morcellation, is in its infancy. No doubt that safe in-bag morcellation of fibroids has the potential to avoid many of the reported morcellation complications, such as direct morcellation injuries, parasitic fibroids and the upstaging of morcellated sarcomas. However, not all risks are addressed such as spillage from the content of the bag in the abdomen especially if the bag is punctured to introduce a laparoscope in the bag for better visualisation of the morcellator’s tip. The in vitro results are promising. Cohen and Einarsson [ 91 ] demonstrated in-bag morcellation in an in vitro study, in which they successfully morcellated beef tongue specimens. In only 1 of the 13 trials did leakage of the bag occur. Washings of the container after retrieval of the specimen bag were negative for muscle cells, except for the open control and the trial with leakage of the retrieval bag. A small series of in-bag morcellation have been published with good results [ 92 , 93 ]. Vaginal morcellation in a bag has been described also [ 94 ]. Possible draw backs are the need for a sufficient vaginal entry and the fact that only hysterectomy specimens are suitable for this technique. Future studies must establish the role of vaginal in-bag morcellation. In urology in-bag (‘contained’) morcellation has been performed in low-grade renal cell carcinoma (RCC). In a retrospective study, Wu et al. [ 95 ] evaluated in-bag morcellation in 188 patients with low-stage RCC. After a mean follow-up of 21 months, no difference in survival was demonstrated compared to open nephrectomy, although one port site metastases occurred. The safety and effectiveness in terms of survival was confirmed in another study [ 96 ]. In low-grade renal cell carcinoma, laparoscopic approach combined with in-bag morcellation of the kidney appears to be safe and effective. Port site metastases after morcellation are thought to be related to the laparoscopic approach itself and not to the morcellation technique as exteriorising the bag before morcellation is meant to prevent contact of the tumour with the abdominal wall. In a review on port site metastases, the in-bag morcellation technique is recommended to prevent port site metastases [ 97 ]. Two studies report on the techniques of containment. Parekh et al. [ 98 ] suggested that specimens should be morcellated in fluid-filled retrieval bags. In an in vitro study, five porcine kidneys were morcellated. No perforation or leakage occurred in the fluid-filled bags compared to one perforation in ‘dry’ morcellation. Also, morcellation time was shorter, when compared to the dry bags. Meng et al. [ 99 ] suggested that washings of a specimen retrieval bag may help the pathological diagnosis after renal morcellation. In their prospective study, 22 consecutive washings were examined. In 9 of 13 patients with carcinoma, the cytological examination confirmed pathology and in three cases cytology provided additional information. In all nine benign cases, cytology was consistent with pathology. The potential benefits and risks of in-bag morcellation should be further evaluated by clinical studies before it can be recommended in general practice. Furthermore, although in-bag morcellation appears to be very promising as a tool for the prevention of morcellation-related complications, further research is needed to improve the morcellation mechanism itself [ 100 ]. The statements on knowledge gaps and potential technical innovations are listed in Table  9 . Table 9 Statements on technical innovation Statements Grade Research on technical innovation in tissue retrieval from the abdominal cavity mainly focusses on in-bag (‘contained’) morcellation D In-bag morcellation may prevent morcellation complications such as direct morcellation injuries, parasitic fibroids and upstaging eventual malignancies Good practice point Potential reported risks of in-bag morcellation is spillage of tumour cells from the bag C In urology in-bag morcellation after laparoscopic removal of early stage and low grade renal cell carcinoma is reported to be safe and effective C Vaginal in-bag morcellation has also been described and needs further study D Development of bags is needed as well as registration of cases to further establish the potential value of on in-bag morcellation in gynaecologic surgery Good practice point Statements on technical innovation

In the previous sections, the results of a literature search are shown to answer questions about fibroid morcellation in laparoscopy, related complications and what is needed to prevent them. It has not been the aim of this study to address morcellation of other structures such as uteri without presumed fibroids (e.g. some cases of LSH) or morcellation of ovarian masses. As expected, the level of evidence is not sufficient to give recommendations. Therefore, it was decided to present options (Table 10 ), which are condensed in a flowchart to offer structure in the clinical management (Fig.  1 ). The flowchart can at best support the clinician to pursue safe and effective fibroid treatment but ‘reassuring’ does not exclude sarcoma and in ‘non-reassuring’ cases the diagnosis will likely be benign. Therefore, it cannot have the status of guideline. Table 10 Options in intended fibroid morcellation Options and considerations Informed consent by the patient is the corner stone of preoperative workup. If fibroid morcellation is intended, include its possible complications in the informed consent procedure before operation. Good practice point Standardise the clinical management by using a flowchart to classify patients according to global risk of a sarcoma in a presumed fibroid. Use flowchart in the figure as an option. Good practice point Use transvaginal ultrasound, transabdominal ultrasound or in case of poor visualisation on ultrasound MRI with or without contrast (Gadolinium-DTPA) D Consider including vascularity parameters (RI and PSV in 2D PowerDoppler ultrasound (PDUS) or vascular indices in 3D PDUS) D Consider performing LDH and iso-enzyme 3 assay D Perform a preoperative endometrial aspiration in case of abnormal uterine bleeding D Support the development of potentially beneficial techniques to prevent morcellation complications by participating in clinical trials Good practice point Register patient’s data after her consent including pre-surgery images and post-surgery histology Good practice point Fig. 1 Flowchart of intended fibroid morcellation Options in intended fibroid morcellation Flowchart of intended fibroid morcellation The ‘additional characteristics’ in the flowchart for women over 40 years of age are based on the literature review in the previous sections. These characteristics have been associated with discrimination of sarcomas from fibroids but they lack thorough scientific evaluation. In the case of LDH, the predictive value is based on only one trial [ 39 ]. We have to realise we are looking for a needle in a haystack indeed. This means that for each single ‘risk factor’, we can easily argue that it is of no importance. However, the association of several risk factors is probably more relevant than a single one. If more risk factors are present, it might be wise to be prudent, although the risk of sarcoma remains low. Fibroid growth has been discussed among the authors at length and it was concluded that ‘growth’ or ‘rapid growth’ are not specifically associated with sarcoma but also with fibroids. On the other hand, a stable size during several months without administration of suppressing hormones (GnRH or Ullipristal) does make a sarcoma highly unlikely. Growth of the presumed fibroids during GnRH treatment or in menopause should raise suspicion about the nature of tumour. The in-bag morcellation of presumed fibroids has the potential to prevent the rare morcellation-related complications typical of ‘open’ morcellation, such as direct morcellation injuries by distending the morcellation bag, the spread of fibroid chips, sarcoma or other malignant particles and cells in the abdominal cavity. More research is needed to improve the technique and on safety before in-bag morcellation can be recommended as a general tool in laparoscopic surgery. We have chosen not to list all the literature on the benefits of laparoscopic surgery which are an important argument in the trade-off between laparoscopy with morcellation and laparotomy without. The vast amount of evidence in favour of laparoscopy is elegantly abstracted in the ‘second look’ article by Pitkin and Parker [ 101 ]. Taking needle biopsies before surgery would be an interesting option in case of intended fibroid surgery to rule out sarcoma; however, there are different opinions on the potential harmful effects of tumour manipulation and spread by needle biopsy [ 45 , 69 , 70 ]. Also representativeness of needle biopsies may be a yet unresolved drawback, given the necrotic areas in the sarcoma. It is clear that there is much research to be done. At first, the issue on sarcoma risk in patients scheduled for fibroid morcellation should be clarified. The preliminary report of Pritts et al. (2014, personal communication) that includes more prospective trials and is based on sound statistics shows much lower prevalences (1:7,400) than previously reported in the meta-analyses based on retrospective trials (1:352) [ 1 ]. Secondly, more data must be collected to estimate the risk of sarcoma in individual patients with presumed fibroids, based on epidemiological data from the patient and diagnostic tests such as imaging. Thirdly, technical innovation, such as in-bag morcellation, should enable safe morcellation of intra-abdominal specimens. A corner stone in the doctor and patient relationship is the informed consent procedure. This means that the patient needs to be informed as completely as possible and based on the best available evidence in order to let her make the best choice of treatment tailored to her individual situation.

This meta-analysis, based largely on peer-reviewed articles and two submitted articles, shows a prevalence of 0.14 % which is lower than but within the range of similar literature reviews [ 1 , 6 ]. In the current selection of studies, the small numbers (<500) were excluded. Recently presented data, submitted for publication, however show that including small trials brings the average reported prevalence of sarcoma down (Pritts E, 2014, personal communication). Pritts et al. selected 131 articles with 29.877 patients operated for fibroids and found a sarcoma prevalence of 1:7,400 (0.014 %). They explain the large prevalence difference with the available literature by including more prospective trials (50 % half of which randomised trials) that lack the confounder of patient selection of retrospective trials. Also trials with smaller numbers were included increasing the power of the meta-analysis using Bayesian statistics to correct for small patient numbers. Although the large differences in prevalence undermine the credibility of all collected data on the prevalence of sarcoma in presumed fibroids, it is likely that the prevalence is much lower than recently reported. Prospective collection of multicentric data of contemporary patients may clarify the important issue of prevalence. As the LMS has a similar clinical and diagnostic appearance to the leiomyoma in contrast to the endometrial stromal sarcoma, our interest is primarily to distinguish fibroids from LMS. The endometrial stromal sarcomas usually cause abnormal—non cyclic—uterine bleeding. But also LMS can present as a type 0—100 % protruding in the uterine cavity—causing abnormal uterine bleeding. In the current review, all kinds of sarcomatous changes were included in the meta-analysis, including STUMP’s. This may have blurred the prevalence results. In particular including studies with laparoscopic supracervical hysterectomies, sometimes without the presumption of fibroids, will result in a lower reported prevalence. A difference in prevalence between studies where fibroids intended to be morcellated and the older (pathology) studies where all uteri and fibroids served as a denominator in the prevalence rate has also been demonstrated [ 27 ]. Data on age and prevalence does not allow the estimation of an accurate risk of sarcoma in the individual patient scheduled for fibroid surgery but they may be taken into account to define a low and intermediate risk group of patients. The paradox remains that as presumed fibroids are less prevalent in postmenopause, the highest (absolute) number of sarcomas is found in the fourth decade, although the incidence is still extremely low [ 23 ]. In one study, 18/21 sarcoma patients (86 %) were premenopausal [ 28 ]. The statements of this section are in Table  2 . Table 2 Statements on the prevalence of uterine sarcoma in presumed fibroids Statements Evidence The incidence of leiomyosarcoma is 0.64/100,000 per year The prevalence of sarcoma in a presumed fibroid is 0.14 % (1:700) with a range from 0.49 % (1:204) to 0.014 % (1:7,400). This large range renders more prospective data collection necessary. C The risk of sarcoma in presumed fibroids is positively related to age, although the majority of sarcomas—in absolute numbers—will be in the fourth decade. Below the age of 40 sarcoma in a presumed fibroid is extremely rare. C Based on age, an accurate assessment of the risk of sarcoma in patients with presumed fibroids is not possible although a global estimation (intermediate risk versus low risk) could be made C Statements on the prevalence of uterine sarcoma in presumed fibroids

A uterine leiomyoma or myoma is a benign smooth muscle tumour of the myometrium and will be referred to in this article as fibroid. On April 17, 2014, the Food and Drug Administration (FDA) published a press release on the website where the use of laparoscopic power morcellation was ‘discouraged’ due to potential upstaging of uterine sarcoma [ 1 ]. The motive to do so was the case history of a patient with a presumed fibroid who underwent a laparoscopic hysterectomy and morcellation of the uterus. The fibroid turned out to be a sarcoma and at reintervention spread of the sarcoma in the abdominal cavity was present. The FDA reported a risk of a uterine sarcoma in patients with presumed fibroids of 0.28 % meta-analysing the data of 18 studies. As some way of morcellation has been used for a long time by gynaecologic surgeons to extract tissue from the vagina or the endoscopic openings in order to enable vaginal or laparoscopic surgery, the professional community was shocked and representatives of many scientific societies published their opinions on the matter [ 2 – 9 ]. A common statement was the lack of solid scientific data to reach strong recommendations with regard to the counselling of patients with fibroids on the issue to choose laparoscopic surgery with its established benefits or for laparotomy to escape the small risks related to morcellation of the fibroid. This review looks at morcellation complications in cases of presumed fibroids. Although intra-abdominal spread of endometrial cancer, adenomyosis and even trophoblastic tissue by morcellation of the uterus has been reported, these issues will not be addressed in this paper as those diagnoses are less unexpected by their clinical features. To explore the available literature, the subject will be divided into clinical questions: What is the risk of sarcoma in patients with a presumed fibroid How to diagnose a uterine sarcoma and distinguish it from a fibroid What are the complications of morcellation How to prevent morcellation complications What are the knowledge gaps Recommendations on clinical management in patients with fibroids What is the risk of sarcoma in patients with a presumed fibroid How to diagnose a uterine sarcoma and distinguish it from a fibroid What are the complications of morcellation How to prevent morcellation complications What are the knowledge gaps Recommendations on clinical management in patients with fibroids The objective of this article is to formulate and grade statements and recommendations on fibroid morcellation based on the level of available evidence. The grading of articles and recommendations was performed according to Eccles et al. [ 10 ] (Table  1 ). As statements can be based on several articles, they are treated as recommendations. Table 1 Grading statements and recommendations [ 10 ] Recommended grade Evidence A Directly based on category I evidence B Directly based on: • Category II evidence, or • Extrapolated recommendation from category I evidence C Directly based on: • Category III evidence, or • Extrapolated recommendation from category I or II evidence D Directly based on: • Category IV evidence, or • Extrapolated recommendation from category I, II or III evidence Good practice point The view of the Guideline Development Group NICE 2002 Recommendation taken from the NICE technology appraisal Evidence category Source Ia Systematic review and meta-analysis of randomised controlled trials Ib At least one randomised controlled trial IIa At least one well-designed controlled study without randomisation IIb At least one other type of well-designed quasi-experimental study III Well-designed non-experimental descriptive studies, such as comparative studies, correlation studies or case studies IV Expert committee reports or opinions and/or clinical experience of respected authorities Adapted from Eccles M, Mason J (2001) How to develop cost-conscious guidelines. Health Technology Assessment 5 (16) Grading statements and recommendations [ 10 ] Directly based on: • Category II evidence, or • Extrapolated recommendation from category I evidence Directly based on: • Category III evidence, or • Extrapolated recommendation from category I or II evidence Directly based on: • Category IV evidence, or • Extrapolated recommendation from category I, II or III evidence Adapted from Eccles M, Mason J (2001) How to develop cost-conscious guidelines. Health Technology Assessment 5 (16)