Introduction

Acute appendicitis is one of the most common surgical emergencies. It is also the most common disorder associated with lawsuits against emergency physicians. The surgical procedure is plagued by negative appendectomies [1]. A negative appendectomy, but also a delay in treatment, can result in unnecessary prolonged hospitalization and morbidity, especially in the elderly [2, 3]. One of the major problems is that multiple disorders can mimic the clinical presentation of appendicitis, especially in women. Traditionally, clinical examination has been used to diagnose appendicitis. However, based on clinical examination only, negative appendectomy rates of 20 % have been reported [4, 5]. It has been suggested that the negative appendectomy rate can drop to as low as 2 % if imaging is added to the diagnostic work-up [1]. Based on these reports, in 2010 the Dutch College of Surgeons published new evidence-based national guidelines concerning the diagnostic work-up of patients suspected of having appendicitis, which stated that an appendectomy should not be carried out without proper preoperative imaging [6]. Ultrasound and computed tomography (CT) have a negative predictive value (NPV) of 72 % and 100 %, respectively, for diagnosing appendicitis [7–9]. Therefore the guidelines stated that a CT scan was compulsory not only after an inconclusive ultrasound, but also after a negative result with ultrasound. However, implementation of these guidelines led to discussion amongst clinicians about whether the extra mandatory imaging would be a cost-effective strategy in reducing the number of unnecessary surgeries.

Therefore the purpose of this study was to evaluate whether the mandatory use of imaging in the diagnostic work-up is an effective strategy in reducing the rate of unnecessary laparotomies in patients with suspected appendicitis. A secondary endpoint of this study was to evaluate in which way this strategy affects the average costs in these patients.

Material and methods

This study was conducted retrospectively. In our country institutional review board (IRB) approval is not required for these types of retrospective studies and patient consent was therefore waived. In our University Hospital all patients are informed that their anonymised data can be used for research purposes. No patient in this study raised an objection to the use of his/her anonymised data.

Patients

The electronic patient databases of the Maastricht University Medical Center were searched for all patients presenting to the surgeon with a suspected acute appendicitis in the differential diagnosis between 1/1/2008 and 12/31/2009 (group I: before guideline implementation) and between 1/1/2011 and 12/31/2012 (group II: after guideline implementation). Patients from the transition period (2010) were not included. Inclusion criteria were: (1) referral by a general practitioner with the suspicion of appendicitis or (2) patients presenting to the emergency department or outpatient clinic with acute pain in the right lower abdomen. Exclusion criteria were: (1) clear alternative clinical diagnosis (e.g., cholecystitis), (2) recent abdominal trauma, and (3) previous appendectomy (which was unknown at the time of patient referral).

Diagnostic work-up before and after guideline implementation

Before implementation of the guidelines there was no standard protocol in our hospital for the diagnostic work-up of patients suspected of having appendicitis. The surgeon on call arbitrarily used clinical (re-)evaluation, diagnostic laparoscopic surgery, CT and/or US for their diagnostic work-up.

The patient routing as described in the national guidelines proposed by The Dutch College of Surgeons in 2010 is displayed in Fig. 1. After implementation of the guidelines the patient routing was as follows: when a patient presents with appendicitis in the differential diagnosis, a surgeon first evaluates the patient. If after clinical and laboratory examination, the surgeon still suspects appendicitis, the patient should proceed to imaging. Ultrasound is recommended as the first-choice imaging technique in patients with suspected appendicitis due to its availability and low costs. Alternatively, CT may also be used as the primary diagnostic tool. In case of a negative or inconclusive ultrasound, the patient should proceed to additional CT examination.

Fig. 1
figure 1

Flowchart showing patient routing derived from the guidelines proposed by the Dutch College of Surgeons. RLQ right lower quadrant

Full size image

Imaging procedures

All ultrasound examinations were performed using an ultrasound machine (Philips Medical Systems, Best, The Netherlands). Ultrasound was performed and evaluated by either a resident (> 1.5 years’ experience in abdominal ultrasound and authorised to perform ultrasound without senior supervision) or a senior radiologist (> 4 years’ experience) on call. CT examinations were performed using multislice CT devices. The standard CT protocol for appendicitis in our institution consisted of a portal venous CT examination through the entire abdomen, performed with a 70-s delay after intravenous administration of 110 cc of Ultravist 300 (Bayer Schering Pharma, Berlin, Germany). Slice thickness was 3 mm and 3-mm coronal reconstructions were routinely constructed. Oral contrast was not standard in the CT protocol for appendicitis. The resident or radiologist on call performed the image evaluation.

Surgical procedures

The standard surgical approach was a laparoscopic appendectomy. When there were complications during the laparoscopic procedure an ‘open’ appendectomy was performed. In case of a normal appendix, the appendix was not routinely removed. The surgeon on call performed the surgery.

Outcome variables and reference standard

A single reviewer (E.M.) analysed each study patient and recorded whether the national guidelines were followed. Additionally, for each patient the reviewer recorded the following parameters: sex, age, type/number of imaging/surgeries, laboratory tests, duration of hospital admission, number of outpatient clinic and emergency room visits, re-admissions, imaging-guided drainage, surgical outcome (normal appendix/appendicitis) and complications during a follow-up of 12 months. Complications were defined as any negative deviation from the normal postoperative course, e.g., abscess, fistula and wound infection. When an appendectomy was performed, histology of the resection specimen served as the reference standard for appendicitis versus a normal appendix. In case of a diagnostic laparoscopy without appendectomy, the surgical assessment of a normal appendix combined with a clinical follow-up of at least 12 months was the standard reference. When no surgery was performed a clinical follow-up of at least 12 months without any further evidence of appendicitis served as the reference standard to confirm the absence of appendicitis.

Cost analysis

Information regarding resource use was collected from the hospital information system for all 1,556 patients during the periods before and after implementation of the guidelines. Cost prices were obtained from the hospital financial department and the Dutch manual for cost research [10].

Since cost data are generally skewed and not normally distributed, a non-parametric bootstrap analysis with 1,000 replications was performed to estimate the confidence intervals surrounding the mean difference in costs [11].

Results

Patients

Baseline patient characteristics are presented in Table 1. In total, 1,556 patients were identified who presented to the emergency department with acute pain in the right lower abdomen and/or who were referred by a general practitioner with a clinical suspicion of appendicitis in the differential diagnosis: 756 before implementation of the guidelines (group I) and 800 after implementation of the guidelines (group II). After clinical examination by a surgeon, 509 (67.3 %) patients in group I and 540 (67.5 %) patients in group II were still suspected of having appendicitis. These patients constituted the final study population for the diagnostic work-up, surgical outcome and assessment of complications. All patients were evaluated for the cost analysis..

Table 1 Patient characteristics
Full size table

Diagnostic work-up and guideline compliance

Details on imaging and guideline compliance are presented in Table 2. In group I, before the guidelines, 289/509 (58.5 %) patients underwent imaging (42 % ultrasound, 12.8 % CT, 3.7 % ultrasound + CT) as part of their diagnostic work-up. In group II, after implementation of the guidelines, 533/540 (98.7 %) of the patients underwent imaging (61.7 % ultrasound, 4.4 % CT, 32.6 % ultrasound + CT, 0.7 % magnetic resonance imaging (MRI)), during their diagnostic work-up. Four MRIs were carried out in group II after implementation of the guidelines (in two pregnant women and two children). No MRI was performed in group I. In 340 of these 540 patients (63 %), the imaging procedures were fully compliant with the guidelines; of the other 190 patients seven patients received no preoperative imaging and in 183 patients additional CT was omitted after an inconclusive (109) or negative (74) ultrasound examination.

Table 2 Overview of surgery and imaging
Full size table

Surgical procedure and outcome

Details on surgical procedures are presented in Table 2. Follow-up of the patients who did not have their appendix removed after surgery showed that none of these patients in either group developed appendicitis. Before the guidelines, 332/509 (65.2 %) patients in group I received surgery, of whom 76 (22.9 %) turned out to have a normal appendix. Fifty-nine of the 76 (78 %) patients with a normal appendix received no preoperative imaging. In group 1, four out of these seventeen patients received a diagnostic work-up similar to the guidelines that were implemented in 2010. After implementation of the guidelines, 274/540 (50.7 %) of the patients in group II received surgery, of whom seventeen (6.2 %) had a normal appendix. Four of these seventeen patients (24 %) did not receive the diagnostic work-up according to the guidelines: in two patients laparoscopy was performed without any preoperative imaging and the other two patients had an inconclusive ultrasound with no additional CT examination. Of the patients with signs of appendicitis on CT images, 3.3 % had no appendicitis at histology. The decrease in the rate of unnecessary surgeries after guideline implementation was significant (p < 0.00001). The surgical complication rate dropped from 19.9 % (66/332) in group I to 14.2 % (17/274) in group II.

Cost analysis

For the cost analysis all 1,556 patients were evaluated (see Table 3). After implementation of the guidelines, the costs of imaging increased significantly on a patient basis from an average of 29 € to 49 €. Conversely, the average duration of hospital stay and costs per patient decreased significantly from 2.75 to 1.91 days, resulting in a drop in costs from 1,335 € to 926 € (confidence level (CL): −763;−67). Average surgical costs decreased from 715 to 551 € per patient (CL: −289;−30). In total, the average costs per patient decreased by 594 € (from 2,482 to 1,888 € (CL: −1081;−143)) after implementation of the guidelines. The average costs per patient with an uncomplicated laparoscopy were 3,939 € in group I and 3,900 € in group II. The average costs per patient with complications were 8,910 € in group I and 8,988 € in group II.

Table 3 Overview of the average number and costs on a patient basis
Full size table

Discussion

The results of our study show that the implementation of the Dutch national clinical practice guidelines for the management of patients with suspected appendicitis recommending mandatory use of imaging led to a significant increase in the use of preoperative imaging. This resulted in a significant reduction in the rate of unnecessary appendectomies from 22.9 % to 6.2 %. Additionally, the overall complication rate after surgery dropped significantly from 19.9 % to 14.4 %. The average costs per patient dropped significantly by 594 € from an average of 2,482 € to an average of 1,888 € per patient, despite a significant increase in the number of imaging examinations.

In patients clinically diagnosed with acute appendicitis, the reported overall negative appendectomy rate (with no routine use of imaging in the diagnostic work-up) is about 15–20 %: 10 % in men and 25–45 % in women of childbearing age [5, 12–14]. This is consistent with the rate of unnecessary laparotomies prior to the guidelines in our hospital (22.9 %) in which almost two-thirds of the patients received no preoperative imaging. Another reason for the relatively high rate of unnecessary laparotomies prior to the guidelines was the common practice in our hospital of performing a diagnostic laparoscopy in cases where there was a strong clinical suspicion of appendicitis.

Overall accuracy of clinical examination for the diagnosis of acute appendicitis is known to be approximately 80 % [15], indicating that in up to 20 % the patients are misdiagnosed. Ultrasound has a positive predictive value (PPV) of 97 % [7]. The PPV of CT is comparable (97 %) with an overall accuracy of (unenhanced) CT up to 98.2 % [16]. This means that theoretically the number of unnecessary laparotomies should be very low with the use of CT in the diagnostic work-up. This was also reported in recent literature, in which very low negative appendectomy rates of 1.7 % were described with the use of preoperative CT imaging [17]. Our data confirm these findings: only 3.3 % of the patients with signs of appendicitis on CT images had a normal appendix and had thus received an unnecessary surgical procedure. Drawbacks of CT imaging like costs and higher radiation dose are becoming less important now that multiple studies have shown that unenhanced low-dose CT can be used to accurately detect/rule out appendicitis [16, 18]. In our study, however, no low-dose CT examinations were made. An additional benefit of CT is that in patients with appendicitis CT imaging accurately demonstrates the full extent of the disease. Furthermore, an alternative cause of abdominal pain may be found with CT in almost one-third of the patients suspected of having appendicitis [19]. Nevertheless, the Dutch national guidelines do not recommend CT as the first-choice imaging modality in the standard work-up. Instead, the guidelines dictate the use of ultrasound as the primary imaging tool. The main reason for this is that, in spite of the moderate negative predictive value, ultrasound has been reported to have a similarly high PPV to that of CT [7]. This means that if appendicitis is diagnosed with ultrasound, the chance of a normal appendix is very low. Furthermore, the costs of ultrasound compared with CT are considerably lower. Ultrasound may, however, be difficult to perform in a few scenarios, such as in patients with severe abdominal pain, patients with overlying intraluminal gas and adipose patients. Furthermore, ultrasound is operator dependant. Therefore the guidelines state that – in case of a ‘contraindication’ for ultrasound, CT imaging can also be used as an alternative primary imaging tool.

Despite the increase in imaging procedures after implementation of the guidelines, we observed a significant decrease in the average costs of care per patient. The main reasons for this reduction in costs was a reduction in the number of laparotomies (from 11.4 % to 6.1 %). This consequently led to a reduction in the average number of hospital admission days (from 2.75 to 1.91 days) and complication rates (from 19.9 % to 14.2 %). The lower complication rate with the use of mandatory imaging may be interpreted as a result of an earlier diagnosis of appendicitis. This plays a key role in cost reduction; the literature shows that early diagnosed, uncomplicated appendicitis carries little morbidity and is relatively inexpensive to treat. However, if the appendicitis progresses, the costs rise exponentially [20]. This was also the case in our study. In addition, selected patients with early-stage appendicitis could benefit from more conservative treatments such as antibiotics [21], lowering the costs even more.

Although the costs of CT compared to a surgical procedure and associated hospital days are very low, our surgeons appeared rather reluctant to order a CT after an inconclusive and in particular after a negative ultrasound examination. Despite scientific evidence, appendicitis is still widely believed to be a ‘simple’ clinical diagnosis [22]. This probably resulted in a suboptimal implementation of the guidelines with consequent lower imaging costs. Theoretically, however, full adherence to the guidelines would still have been a cost-effective approach as the current cost reduction of 594 € per patient means that the addition of CT after inconclusive/negative ultrasound will still result in a reduction in overall costs. Due to the high accuracy of CT it would probably lead to even less unnecessary surgery, resulting in further cost reduction. The suboptimal implementation means that the exact impact of the guidelines could not precisely be defined in this study. Furthermore, the present study was limited due to its retrospective nature. However, the purpose of this study was to evaluate effectiveness of the new guidelines after implementation in a busy daily clinic. Further research is recommended, as a prospective cost-effectiveness study including a social perspective for economic evaluation could help to understand all costs associated with patients suspected of having appendicitis.

In conclusion, this study demonstrates that the implementation of guidelines resulted in a major increase in the use of imaging in the work-up of patients with clinically suspected appendicitis. This resulted in a reduction in negative appendectomies and surgical complications, and reduced the costs of care per patient in the daily clinical practice.