Validation of the Raja Isteri Pengiran Anak Saleha Appendicitis (RIPASA) scoring system for the diagnosis of acute appendicitis among Ethiopian patients: a multi-institutional observational study

Abstract

Background

Acute appendicitis is the most common surgical emergency in Ethiopian clinical practice. Although a multitude of scoring systems have been used in clinical practice, none have been universally validated. The purpose of this study was to validate the Raja Isteri Pengiran Anak Saleha Appendicitis (RIPASA) scoring system in the Ethiopian context.

Methods

A total of 315 consecutive patients who presented with a presumptive diagnosis of acute appendicitis and were planned to undergo appendectomy were studied. All the studied patients had diagnostic sonography and underwent the RIPASA scoring system. The sensitivity, specificity, positive predictive value, and negative predictive value of RIPASA and ultrasound results with intraoperative gross examinations.

Results

The mean age of the participants was 27.4 ± 11.5 years, with a male-to-female ratio of 1.6:1. The concordance between ultrasound and RIPASA for the diagnosis of acute appendicitis was 93.6%. The sensitivity, specificity, positive predictive value, and negative predictive value of RIPASA were 96.2%, 30.8%, 93.9%, and 42.1%, respectively. Similarly, the sensitivity, specificity, positive predictive value, and negative predictive value of ultrasound were 95.3%, 27.8%, 95.6%, and 26.3%, respectively. White cell count and RIPASA scores were weakly correlated with intraoperative stages of acute appendicitis, r(313) = 0.18, p = 0.001, and r(313) = 0.129, p = 0.022, respectively. The rate of a negative appendectomy was 6%.

Conclusion

RIPASA and ultrasound had equivalent performance in the diagnosis of acute appendicitis. In both cases, the rate of negative appendectomy was low enough to validate RIPASA for clinical practice in low-income institutions where sonographic diagnosis by a conventionally trained radiologist is not available.

Background

Globally, the most common surgical emergency is acute appendicitis (AA) [1]. Similarly, data from Ethiopia indicates that appendectomy is the most often performed emergency surgical operation [2]. Although many scoring methods have been created, there is still no clinical diagnostic algorithm or scoring system that is widely accepted, despite the high incidence of AA [3]. The two most often used imaging modalities, computed tomography (CT) and ultrasonography (US), are still not optimal for AA as the rates of false positive and negative cases are non-negligible [4]. Moreover, there are concerns about the accessibility of imaging modalities in low-income environments, particularly in these nations' rural areas [5].

For patients presenting with right iliac fossa (RIF) pain, the Alvarado rating system has long been considered the most favorable scoring method [6]. A new clinical risk scoring system, the Raja Isteri Pengiran Anak Saleha Appendicitis (RIPASA) score, was validated in 2010 in Brunei, Southeast Asia [9]. It comprises 14 clinical parameters, all of which can be readily extracted from a thorough clinical history and physical examination; these parameters and their respective scores are as follows: age (less than 40 years is 1 point; greater than 40 years is 0.5 points), sex (male is 1 point; female is 0.5 points), right iliac fossa (RIF) pain (0.5 points), migration of pain to the RIF (0.5 points), nausea and vomiting (1 point), anorexia (1 point), duration of symptoms (less than 48 h is 1 point; more than 48 h is 0.5 points), RIF tenderness (1 point), guarding (2 points), rebound tenderness (1 point), Rovsing’s sign (2 points), fever (1 point), raised white cell count (1 point), and negative urinalysis (1 point). [7,8,9]. Ever since the RIPASA scoring system was introduced, multiple studies have confirmed that the system performs better than Alvarado scores in various parts of the world [7,8,9]. Furthermore, its sensitivity and specificity were shown to be comparable to those of the US [10]. The comparative sensitivity and specificity of RIPASA with US examination in the Ethiopian setting have not yet been published, despite its validation elsewhere. This study is intended to prove the hypothesis that the RIPASA score is non-inferior to US examination in test accuracy in Ethiopian patients presenting with a presumptive diagnosis of AA when compared to the intraoperative findings.

Methods

Study setting

The study was conducted at two referral hospitals and one specialized hospital in Addis Ababa, the capital of Ethiopia. Within Addis Ababa, there are seven referral hospitals under the city administration and five specialized hospitals under the Federal Ministry of Health. Of these, Zewditu Memorial Hospital and Menelik II Memorial Hospital, from the city administration, and St. Peter Specialized Hospital, from the federal hospitals, are selected. The surgical departments of those hospitals are affiliated with Addis Ababa University. Each month, about 350 and 1050 patients visit Zewditu Memorial Hospital’s emergency and outpatient departments, respectively. Of these, 90 surgical procedures were performed, and for 30 of them, the reason for surgery was appendicitis. On average, 300 and 1000 patients, respectively, visit the emergency and outpatient departments at Menelik II Memorial Hospital each month. Of these, 70 surgical procedures were carried out, 20 of which were performed for AA. Furthermore, based on the review of several months of HMIS at St. Peter Specialized Hospital, on average, 27 appendectomies have been performed.

Study design

A paired study design for diagnostic accuracy has been implemented. The patients were enrolled immediately after they presented with symptoms and signs typical of acute appendicitis, which were confirmed by an ultrasound result performed by a radiologist. All patients were evaluated with a RIPASA scoring system, and all of the patients underwent surgery based on clinical, radiologic, and laboratory results determined by the surgeon’s decision. Intraoperatively, multiple images of the appendix were taken, which were later evaluated by two general surgeons for diagnosis and staging of the appendix along with the surgical team. Postoperative patient follow-up was done until the patients were discharged.

Study population

Source population: All patients who visited the health facility with clinical features of AA at St. Peter Specialized Hospital, Zewditu Memorial Hospital, and Menelik II Memorial Hospital from December 1, 2022, to March 31, 2023.

Study population: All patients who visited the surgical departments of the selected hospitals with a presumed diagnosis of AA within the study period.

Inclusion criteria

Patients over the age of 14 years who presented to the surgical emergency department with a presumptive diagnosis of AA and planned to undergo appendectomy during the study period were included. The decision for appendectomy was not made only by the surgical team based on a composite of clinical, laboratory, and imaging parameters. Patients with inconclusive sonography result or clinical presentations who underwent appendectomy were included in the study.

Exclusion criteria

Excluded patients were those who had undergone an appendectomy, were younger than 14 years old, had a history of chronic abdominal pathology like AA that is considered recurrent during the current presentation, were pregnant and exhibited AA-like clinical symptoms, and had complained of pain for longer than seven days. Patients with comorbid conditions (such as diabetes mellitus, immune-compromised patients, patients on steroids, patients with malignancy, and patients receiving radiation) as well as those with symptoms of generalized peritonitis were also excluded from the study.

Patients who didn’t undergo surgery based on the responsible surgeon’s decision were excluded from the study.

Patients with an age of less than 14 years and pregnant women were excluded because of the paucity of studies in this population on RIPASA. In this situation, a blanket validation study involving all the “at-risk” population along with the adult population can only be done after a separate validation study for the “at-risk” population has proved the scores efficacy. In the Ethiopian health institution’s practice, pediatric age groups below the age of 14 years are treated in pediatric emergency settings, while those that are 14 years old or older are treated along with the adult population.

Variables

Dependent variables

The RIPASA Score outcome, US conclusion, intraoperative gross assessment results, and outcome of the patient after appendectomy.

Independent variables

Sex, age, address, symptom duration, right iliac fossa pain, pain migration to the right iliac fossa, anorexia, nausea and vomiting, right iliac fossa tenderness, guarding, rebound tenderness, rovsing’s sign, fever, white blood cell count, urinalysis, RIPASA score.

Sample size

There has been no single study done in Ethiopia to assess the RIPASA scoring system. A study done in Ireland reported that the diagnostic accuracy of the RIPASA score was 80% at a 95% confidence interval and a 5% margin of error [11]. The sample size was calculated using Epi-Info and is 246. On the other hand, a 10% non-response rate was considered due to incomplete or lost data. The required sample size for the study was 273 participants. Repeated measures from each patient with ultrasound and RIPASA were utilized for data collection, which made a single population proportion sample size calculation a valid method.

At 95% CI, z = 1.96, d = 0.05, p = 0.8, 1-p = q = 0.2, Non response Rate = 0.1 (10%)

$$n=\frac{{{Z}_{\frac{\alpha }{2}}}^{2}P\left(1-P\right)}{{d}^{2}}n=\frac{{\left(1.96\right)}^{2}0.8\left(0.2\right)}{{\left(0.05\right)}^{2}}=246$$

$${n}_{final}=\left(\frac{1}{1-0.1}\right)x246=273$$

$$where,$$

$$n=Sample$$

$$P=the\;proportion\;of\;diagnostic\;accuracy\;of\;RIPASA$$

$$d=Margin\;of\;error$$

$$Z=Standard\;proportion\;population\;at\;95\text{\% }confidence\;interval\left(1.96\right),$$

Sampling procedures

All three hospitals selected–Saint Peter Specialized Hospital, Menelik II, and Zewditu Memorial Hospitals–were selected because of their high number of emergency surgical cases among Addis Ababa hospitals, which was determined after assessing the surgical logbooks of the previous years. All three hospitals contributed one-third of the patient population.

Data collection procedures

Three trained data collectors collected the data using a structured questionnaire that was prepared on the Kobo toolbox, and a separate manual paper was used for a consent form (Annex 1). The data collectors were trained for two days on the objective of the study, which is the validation of the RIPASA scoring system, on how to perform the RIPASA scoring system, and on the use of the Kobo toolbox. The questionnaire had seven parts. Part 1 contains demographic information; Part 2 contains clinical findings on history; Part 3 contains clinical findings on the physical examination; Part 4 contains laboratory results; Part 5 contains the US classification of AA; Part 6 contains intraoperative findings; and Part 7 contains the postoperative outcome (Annex 2).

In order to include all RIPASA score criteria, the US conclusion, and the intraoperative finding categories, the data sources were the patients and their hospital medical records. Since the questionnaire can only be completed by qualified data collectors, the tool was created in English, and verbal interpretation was performed by the data collectors for the patient interview sessions.

The intraoperative findings were assessed by the operating surgeon and then reevaluated by pictures taken intraoperatively by experienced team of surgeons for appendicitis assessment and staging. Postoperative follow-up of the patients was performed until discharge, and the complications mentioned in Annex 1 were recorded at the time of detection during the hospital stay.

The data collection were collected from December 1, 2022, to March 31, 2023. A pretest took place from November 25–30, 2022, at the hospitals where the actual study was intended. A sample size of ten was taken from two of the centers to test the tool. Supervisors, who were part of the study, evaluated quality and completeness of the data.

Operational definition

Acute appendicitis – diagnosed based on the clinical presentation of the patient and confirmed by the intraoperative gross examination by the surgeon as a catarrhal stage appendix, suppurative (phlegmonous) appendix, gangrenous (necrotized) appendix, or perforated appendix. A classification different from the above stages has been considered negative for AA.

Ultrasound criteria for acute appendicitis- a non-compressible tubular right lower quadrant structure with a wall-to-wall diameter of > 6 mm and surrounding inflammatory markings.

Catarrhal Inflammation: The presence of non-purulent fluid in the right lower quadrant or hyperemia of the appendix or non-compressible appendix.

Suppurative (Phlegmonous) Appendix: Dull serosa; dilatation and congestion of surface vessels; fibrinopurulent serosal exudate in well-developed cases; appendix may be increased in diameter and/or dilated appendiceal wall.

Gangrenous (necrotizing) Appendix: Appendiceal wall friable; purple, green, or black.

Perforated Appendix: A hole in the appendix or a fecalith in the abdomen.

Ultrasonographic diagnosis was only accepted when performed by conventionally trained radiologists with a complete written report submitted.

A “conventionally” diagnosed radiologist is a medical doctor who completed medical school training followed by a 3-year training in diagnostic radiology from an accredited residency program and is licensed by the National Health Licensing Authority.

Data management

Data collected from the medical records was entered into an electronic database (Kobo Toolbox) by the data collectors. The completeness of the data was monitored on the electronic database so that it would not pass to the next question unless the required part was not completed. After entry, both advisors from Addis Ababa University verified the data by evaluating the score and the postoperative data. Then the data were exported to Excel and transferred to SPSS version 23. All incomplete data were discarded.

Data analysis procedure

The gathered data were exported to Microsoft Excel, where they were cleaned, coded, and then exported to SPSS version 23 for analysis. Descriptive analysis was used to summarize the patients’ demographic and clinical data using tables, graphs, and charts with frequencies and percentages. A test of homogeneity was performed, and the results were normally distributed (P value > 0.05). A cross-tabulation was used to calculate the specificity, sensitivity, PPV, and NPV of the RIPASA score and US using the intraoperative findings as a reference standard for the diagnostic method. The McNemar test was used to determine the difference in the accuracy of the diagnostic tests. If any difference in the test was not significant or the RIPASA score was favored, the RIPASA score was regarded as non-inferior to the sonography conclusion. If the test accuracy was greater for sonography than for RIPASA, and if the the McNemar test showed statistical significance, RIPASA would be considered inferior. For this study, a preset value for sensitivity and specificity was not assigned because of the significant variability in test accuracy across studies in different centers. Spearman’s correlation was performed to assess the correlation between the white cell count and RIPASA score and the stages of intraoperative appendicitis.

Ethical consideration

The Addis Continental Institute of Public Health, St. Peter Specialized Hospital, Addis Ababa Public Health Research and Emergency Management Directorate, and the corresponding hospitals all gave their ethical approval before the study could be carried out. We obtained permission from each hospital's management to view the patients' medical records, while interviewing the patients. The questionnaire was strictly followed to gather the data. All patients signed written informed consent form if they were 18 years or older, and for those who were younger, assent was obtained from the patients while written consent was signed by the legal guardians. Patient information is kept private. The fundamentals of research ethics were followed in the conduct of the study.

The study was conducted in accordance with the Helsinki Declaration and National and Institutional guidelines.

Results

A total of 315 patients fulfilled the inclusion criteria and were included in the study. The participant flow diagram representing the patient’s presentation, recruitment, management, follow-up, and discharge status is presented in supplemental 1. The mean age of the participants was 27.4 ± 11.5 years (range: 14–80 years), and there was a 1.6-fold male predominance. Nearly two-thirds of the patients presented within 48 h of symptom onset. Table 1

Table 1 Sociodemographic profile of the study participants

RIF pain and iliac fossa tenderness were reported by 99% and 98.4% of the patients, respectively. The mean white blood cell count (WCC) at presentation was 12918.3 ± 4766.6 cells/mm³ (range: 1300 – 27,500cells/mm.³), and 73.7% had an elevated WCC above the cutoff value. Only 4 patients had a positive urinalysis result. Guarding was the least common clinical sign. Table 2

Table 2 Clinical presentation of the participants

The overall rate of negative appendectomy was 19 (6%) based on the intraoperative findings. The most common stages of appendicitis during appendectomy were the suppurative/phlegmonous (44.1%) and catarrhal (22.9%) stages. Figure 1

Fig. 1

Stages of acute appendicitis

All patients underwent an abdominal sonographic study. A sonographic diagnosis of AA was made in 297 (94.3%) patients, while normal appendix and non-visualized appendix results were reported in 10 and 8 patients, respectively. When correlated with the intraoperative result, the sensitivity, specificity, positive predictive value, and negative predictive value of sonography were 95.3%, 26.3%, 95.6%, and 27.8%, respectively. US missed a total of 13 appendicitis cases and mislabeled 14 negative appendectomies. Table 3

Table 3 Comparison of intraoperative, RIPASA and US diagnoses

The sensitivity, specificity, positive predictive value, and negative predictive value of the RIPASA score were 96.2%, 42.1%, 93.9%, and 30.8%, respectively. A total of 18 and 11 true appendicitis cases and negative appendix cases, respectively, were mislabeled using this scoring system. US and RIPASA scores were concordant in 93.6% of the patients. McNemar’s exact test revealed that there was a statistically significant difference between the two diagnostic tests favoring RIPASA (p = 0.047).

Two weak positive correlations were found between the WCC, RIPASA score, and intraoperative stage of appendicitis. The WCC showed a positive correlation with the stage of appendicitis, r(313) = 0.18, p = 0.001. Similarly, the RIPASA score correlated with the stages of appendicitis r(313) = 0.129, p = 0.022.

Only three complications were detected among the patients studied during the in-hospital course of the patient follow-up: one respiratory infectious complication and two wound infections. There was no mortality recorded during the course of the study.

Discussion

Compared with intraoperative evaluation, both the RIPASA and US exhibited concordance rates of almost 94% and equivalent sensitivities and PPV in terms of diagnostic accuracy. The results of both tests showed low NPV and specificities, while the RIPASA had a small but significantly greater score in the McNemar test. Although the correlation was weak, the only individual symptoms that were substantially correlated with the diagnosis of AA were anorexia, nausea/vomiting, and migrating right lower quadrant discomfort. This investigation verified that, for diagnosing AA, the RIPASA score at least equaled the US performance.

The Alvarado score, first developed by Alfredo Alvarado in 1986 based on a retrospective study including 305 patients, was the first widely used scoring system [12]. Subsequent reports of Alvarado score modifications indicate a moderate improvement in diagnostic accuracy compared to the original Alvarado score [13, 14]. According to reports, the accuracy scores of other scoring systems, such as the adult appendicitis score (AAS), Lintula, and appendicitis inflammatory response (AIR), are inconsistent and somewhat equal to those of the Alvarado score [15,16,17]. It is noteworthy that although the diagnosis relied solely on clinical and US findings, with or without the use of a CT scan, the Jerusalem guidelines of the World Society of Emergency Surgery recognized the AAS and AIR as clinical predictors [18]. Compared to existing scoring systems such as the AAS and AIR, the RIPASA, a relatively new scoring method, has continuously demonstrated a high degree of sensitivity and specificity [7,8,9, 16]. The benefit of RIPASA over the AAS and AIR is that it does not depend on inflammatory markers or differential polymorphonuclear counts for risk prediction, which are tests that are difficult to get in low-income institutions [16, 19].

Because US is operator-dependent, sonography conducted by radiologists with conventional training plays a critical role in the diagnosis of AA [20]. It is logistically and financially difficult to rely too heavily on the US for diagnosis in low-income nations like Ethiopia, where the ratio of radiologists to population is greater than 1 to 350,000 [21]. If the cutoff values are at least as reliable as the US, clinical scoring systems may be able to lessen the reliance on the US for the bulk of patient diagnoses of AA. We were able to verify that RIPASA is at least as accurate and sensitive as the US reports with our study.

It is worthwhile to discuss the low specificity of both diagnostic techniques. Merely nineteen of the patients that underwent surgery were found to have no inflammatory signs associated with AA. Eight patients could have avoided needless appendectomies if the RIPASA score had been used. One further thing to take into account is that this study was carried out in referral centers, where all cases with negative US and clinical findings of AA were not referred, and those with unclear and diagnostic results that needed a surgeon's assessment were transferred to referral centers. It's possible that this decreased both tests' specificity. However, both tests yielded results that were sufficiently good, with the negative appendectomy rate falling within the range that would be considered acceptable in a typical clinical setting [22].

The strengths of this study were the large sample size and the prospective nature of the study. In addition, we utilized only US results that were reported by conventionally trained radiologists. Trained data collectors performed the RIPASA scoring for all the patients, and intraoperative data were evaluated using a team of experienced surgeons to confirm all intraoperative diagnoses of appendicitis.

The weaknesses of this study included the use of a gross examination of the appendix over the histopathologic examination, which was logistically impossible in the institutions of study. Nonetheless, close to 80% of the patients had complicated acute appendicitis, with suppurative/phlegmonous, perforated, and gangrenous appendicitis making the visual diagnosis of AA sensitive.

In conclusion, RIPASA and US had a high concordance rate, good sensitivity, and a positive predictive value in comparison to intraoperative findings. The rate of negative appendectomies was low enough in both to allow for clinical utilization of RIPASA in facilities with low-income institutions with low access to US examinations by conventionally trained personnel.