Management of acute testicular pain in children: changing trends with improvements in scrotal ultrasonography over 18 years

Abstract

Introduction:

Doppler scrotal ultrasonography (US) is the modality of choice in diagnosing testicular torsion. We aimed to evaluate the performance of scrotal US in diagnosing testicular torsion over the past 18 years in our institution and determine the factors contributing to the length of wait times for it.

Methods:

A retrospective review was conducted of boys who presented with acute scrotal pain from 2014 to 2015. US reports, operative findings, final diagnosis and key time points of the patients’ journey (time to emergency department consultation, time to admission, time to US and time to operating theatre [OT]) were collected. US performance results were compared with those observed in a historical cohort from 1998 to 2004. Wait times were compared between operated and non-operated patients.

Results:

Data from 519 boys with a mean age of 9.15 years was collected. Of these, 438 (84.4%) boys had undergone initial scrotal US; of these scrotal US cases, 28 were surgically explored, with 23 confirmed to have torsion. Another five cases were explored without prior US, and all were confirmed to have torsion. Performance analysis of US showed a sensitivity of 100% and a specificity of 98.8%. There was no significant difference between wait times of operated and non-operated patients. Time to US (P < 0.0001, r = 0.96) and time to OT (P < 0.0001, r = 0.64) correlated significantly with the total time from presentation to surgery.

Conclusion:

There has been an improvement in the diagnostic performance of scrotal US for testicular torsion over the past 18 years. Quality improvement programmes targeted at reducing wait times for patients presenting with acute scrotum should target time to US and time to OT.

INTRODUCTION

Assessment of acute scrotal pain in children and adolescents remains a challenge because of its myriad causes. It is universally acknowledged that testicular torsion is a true surgical emergency requiring urgent scrotal exploration.[1]

In 2005, a paper published by our institution reviewed children with acute scrotum presenting between 1998 and 2004, and reported on the addition of colour Doppler ultrasonography (US) to clinical assessment of patients with acute scrotum and its impact on diagnostic accuracy.[2] The authors concluded that the number of explorations and length of hospital stay were reduced along with the cost of management when Doppler US was used. Thereafter, most children in our institution who present with acute scrotum undergo US of the scrotum as part of their clinical assessment.

The primary aim of our study was to evaluate the practice trends and diagnostic performance of scrotal US in the assessment of acute scrotal pain in children, comparing a contemporary cohort with our previously reported historical cohort. Our secondary aim was to examine the key intrahospital patient journey time points to determine the factors contributing to wait times. Our hypothesis is that following the introduction of US as a part of routine assessment, the accuracy of US has improved, leading to lower rates of negative exploration in our institution.

METHODS

After receiving local institutional ethical approval (no. 2016/2328), we conducted a retrospective study looking at all admission data from 1 January 2014 to 31 December 2015 of all children who presented to the emergency department (ED) of KK Women's and Children's Hospital with a complaint of testicular or scrotal pain by reviewing the hospital's electronic medical records (EMR). We excluded boys who were discharged directly from the ED without admission, and those who had a history of direct testicular trauma. We collected demographic data and details of inpatient management, including US reports, operative findings and the final diagnosis. Data between this study's cohort and our previous historical cohort (1998–2004) was compared.

Additional data was collected to analyse the various intrahospital key time points of patients presenting with acute scrotum. These included duration of symptoms, time of triage, ED consultation, admission, US and surgery. Duration of symptoms was defined as the duration of the presenting complaint as documented by the ED physicians in their clinical notes; time of triage was the time stamp on the patient's EMR indicating the time of the first triage in ED; time of ED consultation was represented by the time stamp on the patient's EMR when first reviewed by a physician after triage; time of admission referred to the time stamp on the patient's EMR indicating the time of admission to the surgical inpatient ward; time of US was indicated by the time stamp on the first US image obtained from the scrotal US (if US was performed); and time of surgery was indicated by the time stamp on the patient's operative notes, indicating the time when the patient was wheeled into the operating theatre (OT).

In our institution, patients with acute scrotal pain enter an expedited triage and admissions pathway from the ED. Point-of-care ultrasonography (POCUS) services are not available in the ED, and imaging can be performed only when the patient is registered as an inpatient. Although most patients with scrotal pain do undergo US, this is subject to clinical assessment by the attending surgeon. Where clinical suspicion is high, the patient may be advised to undergo immediate surgery without preoperative imaging. Those who do not undergo scrotal exploration are kept for inpatient observation for a minimum of 24 hours to ensure non-progression of symptoms. After discharge, all patients, including those who do not undergo inpatient US scans, receive a follow-up telephone call 1–2 days after discharge to confirm resolution of symptoms.

All US scans were performed by either a radiology consultant or a radiology resident with consultant supervision. Findings suggestive of torsion included any or all the following: (a) a spermatic-cord whirlpool sign; (b) reduction and/or absence of venous and/or arterial signal within the testicle; (c) heterogeneity of testicular echotexture; (d) horizontal or altered lie; and (e) globular testicular enlargement. US examinations were performed using a 5- to 12-MHz linear transducer (EPIQ 7G and iU22; Philips Medical Systems, Bothell, WA, USA).

We used chi-square tests for categorical variables and Mann-Whitney U tests for nonparametric continuous variables to compare the groups. The diagnostic accuracy of US was calculated using sensitivity, specificity and predictive values (positive and negative). Correlation analysis was performed to identify the relationship of wait times with overall time to surgery for operated cases. A P value <0.05 was considered significant, and IBM SPSS Statistics version 23.0 (IBM Corp, Armonk, NY, USA) was used for statistical analysis.

RESULTS

In total, 722 boys presented to the ED with scrotal pain from 2014 to 2015. Of these, we excluded 175 boys who were treated and discharged from the ED without requiring inpatient admission, 19 who had a history of direct testicular trauma, and nine who had incomplete data. This left 519 patients for inclusion, with a median age of 9.8 years (interquartile range = 5.06 years), of whom 438 (84.4%) had undergone initial scrotal US and 81 had not. Table 1 details the diagnoses of all patients at discharge. We found that torsion of an epididymal cyst was a common diagnosis in our series, which is consistent with other reports.[3]

Table 1.

Diagnoses of patients at discharge.

Diagnosis at discharge	No. of patients
Underwent ultrasonography
Torsion of epididymal cyst	175
Epididymitis/epididymo-orchitis	126
Nonspecific testicular/scrotal pain	72
Testicular torsion	23
Hydrocele	17
Varicocele	4
Others*	21
Did not undergo ultrasonography
Torsion of epididymal cyst	37
Nonspecific testicular/scrotal pain	15
Testicular torsion	5
Hydrocele	5
Inguinal hernia	1
Varicocele	1
Others*	17

*Others includes discharge diagnoses such as viral illness, gastroenteritis, dermatitis and cellulitis.

A total of 438 patients were included in our analysis of the diagnostic performance of scrotal US [Figure 1]. Of these, 28 (6.4%) underwent surgical exploration based on initial US impression; of these, 23 (82.1%) were confirmed to have testicular torsion. All the remaining 410 patients, who had initial negative US reports for testicular torsion and were managed conservatively, were discharged after a period of observation without readmission for testicular torsion. Routine follow-up telephone calls to all discharged patients also revealed no missed cases of torsion. Of the 81 (15.6%) patients who did not undergo US, five were surgically explored on the basis of clinical suspicion, and all were confirmed to have testicular torsion (negative exploration rate = 0%).

Figure 1.

Flowchart shows the study cohort.

When we compared the diagnostic performance of scrotal US in our study cohort with that of our historical cohort, we observed improvements in sensitivity and negative predictive value [Table 2]. However, there were five false positives in our study cohort, resulting in lower specificity and positive predictive value. There was no significant difference in long-term outcomes between patients who initially underwent US and those who did not (P = 0.57). Unfortunately, we were unable to accurately determine the concordance rate between clinical diagnosis and diagnosis made by US, on the basis of our chart review.

Table 2.

Comparison of diagnostic performance of scrotal US for acute scrotal pain and clinical outcomes between the present study cohort and a historical cohort.a

Variable	n
	Current study, 2014–2015 (n=438)	Historical cohort, 1998–2004 (n=332)
Diagnostic performance of US
False positive	5	0
False negative	0	4
True positive	23	9
True negative	412	319
Negative predictive value (%)	100.0	98.8
Positive predictive value (%)	82.1	100.0
Sensitivity (%)	100.0	69.2
Specificity (%)	98.8	100.0
Clinical outcomes
No. of testes lostb	10	22
No. of cases of testicular torsion	28	36
Testes loss (%)	35.7	61.1
Negative exploration rate (%)	9.7	43.3

^aData from Lam et al.[2] ^bAt the time of surgery (i.e. an orchidectomy was performed). US: ultrasonography

No patient who was initially discharged was readmitted or reattended with scrotal pain. Of the patients who underwent surgical exploration and testicular salvage, 5 (35.7%) out of 14 who initially underwent US had subsequent atrophy (mean follow-up period = 13.9 months) and 1 (25.0%) out of four who initially did not undergo US had subsequent atrophy (mean follow-up period = 25.3 months).

The total time from triage to OT for operated patients was 4.61 (range 1.87–18.58) hours. A breakdown of various patient time points can be found in Table 3. There were no significant differences in key wait times between the operated and non-operated cohort [Table 3]. However, when we reviewed only the operated cases, we found that the wait time for US (P < 0.0001, r = 0.96) and wait time for surgery (P < 0.0001, r = 0.64) correlated significantly with the total wait time, whereas the wait time for ED consultation (P = 0.24, r = 0.22) and wait time for admission (P = 0.28, r = 0.20) did not [Figure 2]. When we reviewed the duration of symptoms, the documentation was inconsistent in terms of the amount and type of details provided. The qualitative descriptions also varied. For example, ‘scrotal swelling’ was sometimes recorded as the presenting complaint rather than ‘pain’. Therefore, we could not perform a meaningful analysis on the effect of symptom duration on outcomes.

Table 3.

Key intrahospital wait times for operated vs. non-operated cases.

Timings	Median (range)		P
	Operated	Non-operated
Time from triage to ED consultation (hr)	0.45 (0.01–1.48)	0.40 (0.08–4.25)	0.439
Time from ED consultation to admission (hr)	0.52 (0.34–1.78)	0.61 (0.05–11.58)	0.396
Time from admission to scrotal ultrasonography (hr)	3.41 (2.12–13.18)	3.75 (0.85–23.48)	0.439
Time from scrotal ultrasonography OR admission to surgerya (hr)	4.50 (1.87–18.58)	NA	NA
Total time from triage to OT (hr)	4.61 (1.87–18.58)	NA	NA

^aFive operated patients did not undergo scrotal ultrasonography. ED: emergency department, NA: not applicable, OT: operating theatre

Figure 2.

Graph shows the correlation between the total time from triage to operating theatre (OT) and various key wait times. (a) Time from triage to emergency department (ED) consultation vs. total time from triage to OT. (b) Time from ED consultation to admission vs. total time from triage to OT. (c) Time from admission to scrotal ultrasonography vs. total time from triage to OT. (d) Time to surgery from scrotal ultrasonography compared to total time from triage to OT.

DISCUSSION

Our initial hypothesis that the improvement of accuracy of US over the years has led to decreased negative exploration rates is demonstrated in our study, where we observed improvements in sensitivity and negative predictive value over time. The negative exploration rate also decreased dramatically from 43.3% to 9.7%. However, although US reduces the number of negative scrotal explorations, the wait time for US contributes to the overall wait time and should be targeted for improvement in management pathways for acute scrotum in our institution.

If a torted testicle is treated within six hours of the presenting pain, there is a good chance of saving it in 90%–100% of cases. If treated within 6–12 hours, depending on the degree of torsion, 20%–50% of testicles can be saved, and if treated within 12–24 hours, only 0%–10% of testicles can be saved.[4] Nevertheless, the symptoms and signs often overlap with other causes of scrotal pain. The traditional practice of immediate surgery for all suspected cases of torsion can result in unnecessary operative procedures in up to 92% of cases.[2] US with colour Doppler as an imaging modality has been shown to be highly specific and sensitive with regard to the identification of testicular torsion compared with other causes of acute scrotal pain, reducing the number of unnecessary surgical explorations and quickly identifying patients with true testicular torsion.[2,5] However, detractors caution that waiting for US causes crucial delays in a time-sensitive emergency and that US may be falsely negative.

In our practice, scrotal exploration dropped from 92% in 1998 to 6.4% in our contemporary cohort. Along with this shift, there was a reduction in the proportion of testes lost. We also observed an improvement in sensitivity and negative predictive value, likely reflecting a combination of the following factors: (a) advances in machine technology specifications, allowing greater detail in the obtained US images; (b) enhanced evidence-based protocols for diagnosing acute scrotal conditions on US imaging[6,7,8]; and (c) improvement in operator skills and experience because of the volume of performed scans. There were three cases of false positives, accounting for the slight decrease in specificity, which we believe to be an acceptable result of the reporting radiologist erring on the side of over- rather than under-diagnosis.

Despite many studies documenting the usefulness of scrotal US in the evaluation of acute testicular pain, its role in routine management still remains controversial. The main argument centres on the time-sensitive nature of testicular torsion and the impact on testicular salvage.[9,10,11] However, the actual correlation between testicular salvage rate and symptom duration still varies in the literature. Some argue that US itself causes delays in the diagnosis and management of testicular torsion. However, studies by Arevalo et al.[12] and Afsarlar et al.,[13] which compared groups of paediatric patients with testicular torsion with accelerated and normal workflows, found no significant difference between the groups in terms of orchidectomy rates. These studies propose the idea that the viability of the testis in cases of torsion is predetermined, that is determined before the patient arrives at the hospital. Our study looked at the key intrahospital time points and showed that the wait time for scrotal US indeed played a significant role in patients’ total wait time (P < 0.0001, r = 0.96), but we could not analyse the effects of symptom duration. By contrast, there are clear advantages of US in reducing the negative exploration rate, reducing costs and sparing both patients and caregivers surgical risk.[14] However, we do acknowledge that the delay caused by scrotal US wait time may not be the same in every hospital system.

One possible way to reduce the wait time for patients with an acute scrotum is POCUS performed by a trained emergency physician. The efficacy of POCUS has been demonstrated and adapted in areas such as Focused Assessment with Sonography for Trauma scans, goal-directed echocardiography, US in cardiac arrest and thoracic US.[15] The few studies that have been published on POCUS in acute scrotum report high levels of sensitivity and specificity of up to 95% and 96%, respectively.[16,17] However, these studies were not conducted in paediatric patients, where we could expect reduced accuracy owing to factors such as smaller testicular and vascular dimensions, and lack of cooperation in a child with pain.

Reasons for delayed US diagnosis include the operator-dependent nature of US and difficulty in visualising and interpreting the signs that point to torsion.[18] Other imaging modalities have been explored to address the disadvantages of US, including non-invasive optical techniques such as near-infrared spectroscopy, magnetic resonance imaging and radionuclide scans.[19,20,21,22,23,24] These have remained either experimental, with few or no studies carried out in children, or have limited use in clinical settings because of logistic and technological requirements. The use of a scoring system might help stratify the risk so that patients with highly positive scores are sent directly to the OT.[25]

We acknowledge the limitations of our study, which are inherent to its retrospective design. We could not evaluate the impact of symptom duration because of inconsistent formats in documentation. Electronic time stamps, while providing clarity, may not accurately reflect the timing of the actual event. In addition, this study was conducted using clinical data of a single institution that operates within one standard workflow. Given that wait times are highly dependent on workflow, the data may not be representative of every institution. Despite these drawbacks, our study involves a large sample size, and we were able to provide an overview of practice trends spanning 18 years.

In conclusion, scrotal US has shown consistently good results in sensitivity and negative predictive value over the past decades. We found that US has led to increases in wait time for patients presenting with testicular pain in our institution. Although there are currently no clear criteria for determining the necessity of scrotal US, quality improvement projects should aim to reduce wait times for both US and surgery while we await new evidence.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

About the first author

Dr Lin Kyaw is a Medical Officer who graduated from the National University of Singapore, Yong Loo Lin School of Medicine, in 2020. As a junior doctor, he strives to integrate evidence-based medicine in his day to day clinical work. He also has a keen interest in research and hopes to contribute academic work in the field of Urology.