The study, led by Anjali R. Magan, Nomasonto N. Mkhize and Maureen M. Ramantsi of the Department of Diagnostic Radiology at the University of the Witwatersrand, was designed as a retrospective, single-centre descriptive audit. Rather than tracking every procedure continuously over five years, the researchers sampled distinct audit windows: one contiguous six-month pre-pandemic sample running from October 2018 to April 2019, three discrete two-month windows nested within the first, second and third pandemic waves of 2020 and 2021, and one contiguous six-month later audit-period sample from October 2022 to April 2023. In total, 1,443 procedure records were analysed. The authors are careful to stress that the intervening periods were not sampled, meaning the audit describes the windows examined rather than providing an unbroken chronicle of the service.

The headline finding is strikingly consistent: drainage and decompression dominate the workload. Pigtail catheter drainage and aspiration, a technique in which a coiled-tip catheter is placed percutaneously into abscesses, effusions or other fluid collections to evacuate their contents, accounted for 566 of the 1,443 records, or 39.2 percent of all recorded activity. Biliary drainage interventions, including percutaneous transhepatic biliary drainage and percutaneous transhepatic cholangiography, formed the second-largest group at 282 records, or 19.5 percent. Percutaneous nephrostomy, the image-guided placement of a tube directly into the kidney to bypass obstruction of the urinary tract, contributed a further 128 records, or 8.9 percent. Together, these three categories illustrate a service whose core business is relieving pressure and removing fluid, often in patients too unwell for surgical intervention.

The workload distribution across the sampled periods tells its own story. The pre-pandemic sample contained 601 records, the pooled pandemic-wave sample 387 records, and the later audit-period sample 455 records. Within the pandemic waves themselves, the three two-month windows contained 105, 115 and 167 records respectively, suggesting a gradual recovery in recorded activity as the waves progressed. The drop from the pre-pandemic sample to the pooled wave sample is consistent with the enormous disruption that COVID-19 imposed on elective and semi-elective procedures worldwide, as theatres were repurposed, staff were redeployed and patients delayed presentation. Yet the authors explicitly caution against causal interpretation: because the audit comprised discrete windows rather than continuous observation, the findings describe the sampled periods and cannot establish that the pandemic itself drove the changes.

To test whether the composition of the procedure mix shifted over time, the researchers turned to exploratory, procedure-record-level Pearson chi-square tests, comparing seven analytical procedure categories across the three broad samples and separately across the three wave windows. For the broad samples, the test yielded a chi-square statistic of 41.13 with 12 degrees of freedom and a p value below 0.001, with Cramér’s V, a standardised measure of association, at 0.119. For the wave windows, the chi-square was 32.08 with 12 degrees of freedom and a p value of 0.0013, with Cramér’s V of 0.204. In plain terms, the procedure mix varied significantly across the sampled intervals, and the effect size was modest for the broad comparison but moderate for the wave-to-wave comparison. These statistics are exploratory by design, and the authors flag an important limitation: because persistent patient identifiers were unavailable, the p values do not account for possible within-patient clustering, meaning the same patient contributing multiple procedures could influence the results.

Why does a single-centre audit from a Johannesburg hospital matter to the wider world? The answer lies in the scarcity of site-specific data on interventional radiology activity in South African public hospitals. Health service planning, whether for imaging capacity, specialist staffing, consumables procurement or training programmes, depends on knowing what procedures are actually performed, in what volumes, and how demand fluctuates under stress. Chris Hani Baragwanath Academic Hospital serves an enormous and diverse population, and its interventional radiology service operates within the resource constraints typical of the public sector, where equipment downtime, contrast shortages and competing clinical priorities can reshape a service overnight. A benchmark derived from 1,443 real procedure records gives planners and clinicians something they have rarely had: an evidence-based picture of demand.

The technical profile of the service also carries implications for training. Pigtail drainage, biliary intervention and nephrostomy are bread-and-butter procedures in many interventional radiology curricula, but the balance of categories observed at CHBAH reflects the specific disease burden of its catchment population, including high rates of infective and obstructive pathology. Trainees rotating through such a service encounter a case mix that differs from private-practice or high-income-country patterns, where oncological embolisation, vascular intervention and tumour ablation often dominate. Understanding this profile helps the University of the Witwatersrand align its fellowship and registrar training with the realities of the service, and helps other institutions in similar settings anticipate the skills and equipment their own services will require.

The pandemic windows add a layer of historical significance. Waves I, II and III of COVID-19 in South Africa, driven in turn by ancestral strains, the Beta variant and the Delta variant, placed unprecedented strain on hospital infrastructure. Imaging departments were central to that response, providing chest radiography and computed tomography at scale while maintaining, where possible, the interventional procedures that patients with abscesses, biliary obstruction and renal colic could not safely defer. The rising record counts across the three wave windows, from 105 to 115 to 167, hint at a service adapting under pressure, though the sampled design means the authors stop short of characterising this as a trend. What the audit does establish is that interventional radiology remained active throughout the pandemic periods examined, a testament to the essential nature of image-guided drainage and decompression.

Methodologically, the study is transparent about its limits. It is a Level 4 retrospective descriptive study, conducted in accordance with the Declaration of Helsinki and approved by the Human Research Ethics Committee of the University of the Witwatersrand, which waived individual informed consent given the retrospective design. The sampling strategy trades comprehensiveness for feasibility, and the absence of patient-level linkage constrains the statistical inference. Yet these constraints are honestly declared, and the descriptive benchmark the audit provides is precisely the kind of foundational data that prospective monitoring programmes can build upon. The authors frame their work as a service-level benchmark for planning imaging capacity, specialist staffing, consumables and training, and for prospective monitoring of demand, access and outcomes, a mandate that extends well beyond a single institution.

As health systems worldwide digest the lessons of the pandemic, studies like this one demonstrate the value of auditing what actually happens at the bedside, or rather at the angiography table. The picture that emerges from Chris Hani Baragwanath Academic Hospital is of an interventional radiology service anchored by drainage and decompression work, responsive to shifting demand across turbulent years, and now equipped with baseline data to argue for the capacity it needs. For a field that often measures itself by technological spectacle, the humble pigtail catheter, placed under imaging guidance to drain an abscess or decompress a blocked kidney, turns out to be the quiet workhorse of one of the world’s largest hospitals, and now one of its best-documented.

Subject of Research: Interventional radiology service audit at a large South African public hospital

Article Title: Body interventional radiology procedure patterns across selected audit periods, 2018–2023, at Chris Hani Baragwanath Academic Hospital, South Africa: a descriptive service audit

Article References: Body interventional radiology procedure patterns across selected audit periods, 2018–2023, at Chris Hani Baragwanath Academic Hospital, South Africa: a descriptive service audit. (n.d.). https://doi.org/10.1186/s12880-026-02904-0

Image Credits: AI Generated

DOI: 10.1186/s12880-026-02904-0

Keywords: interventional radiology, Chris Hani Baragwanath Academic Hospital, South Africa, COVID-19, procedure audit, pigtail catheter drainage, biliary drainage, percutaneous nephrostomy, tertiary hospital, BMC Medical Imaging, service planning, public healthcare

The research, conducted by Bowen Yang and Long Jin of the hospital’s Department of Interventional Radiology, took a deliberately multifaceted approach to the problem. Rather than relying on a single dosimetric number, the team analyzed three complementary radiation metrics in 100 patients who underwent PTCD. The first was fluoroscopy time, the total duration of live X-ray imaging during the procedure. The second was cumulative dose, which reflects the intensity of the radiation beam integrated over the exposure time. The third was dose-area product, which combines the cumulative dose with the size of the irradiated field, capturing how much tissue is exposed as well as how strongly. Each metric tells a different part of the story, and the researchers argued that only by examining all three together can clinicians understand where radiation risk truly originates.

The statistical strategy behind the study is worth appreciating, because it addresses a subtle trap in radiation research. Fluoroscopy time, cumulative dose, and dose-area product are not independent quantities. Cumulative dose is, by definition, dose rate multiplied by exposure time, and dose-area product is cumulative dose multiplied by the irradiated field area. If a researcher simply throws all three into a regression model alongside patient characteristics, the results become difficult to interpret, since the dose metrics are physically entangled with one another. Yang and Jin solved this by building a chain of multivariate generalized linear models using a Gamma distribution with a log link, a choice well suited to skewed, strictly positive dosimetric data. Fluoroscopy time was retained as a covariate in the cumulative dose model, and both fluoroscopy time and cumulative dose were retained in the dose-area product model. This allowed the team to partition the intrinsic physical contribution of exposure duration and dose intensity, so that any remaining significant factors could be interpreted as influences on each dose component beyond those mechanical relationships.

The findings on fluoroscopy time were among the most striking. Three factors emerged as significant predictors. Patients in whom the drainage was accessed via a subxiphoid route, rather than a dual approach, had substantially shorter fluoroscopy times, with the model estimating the subxiphoid route at roughly a third of the dual route’s exposure duration. Obstruction located at the hepatic hilum, as opposed to the common bile duct, was also associated with shorter fluoroscopy time, a somewhat counterintuitive result that may reflect differences in how easily guidewires and catheters can be maneuvered depending on where the blockage sits. Finally, older age predicted shorter fluoroscopy times, with each additional year of age corresponding to a small but statistically significant reduction. Taken together, these results suggest that the geometry of the access path and the location of the obstruction shape how long operators must work under continuous imaging.

When the researchers turned to cumulative dose after adjusting for fluoroscopy time, a different set of drivers came into focus. The dominant predictor was body mass index. Heavier patients absorbed markedly more radiation, with each unit increase in BMI associated with a roughly nine percent increase in cumulative dose. This makes physical sense: more tissue between the X-ray source and the detector forces the machine to increase output to penetrate the body, raising the dose delivered for every second of imaging. Female sex was associated with lower cumulative dose, and interestingly, patients with less severe intrahepatic ductal dilatation also received lower doses than those with severe dilatation, an association that reached statistical significance. Longer fluoroscopy time, unsurprisingly, remained a strong predictor of higher cumulative dose, confirming that duration and dose intensity both matter independently.

The dose-area product model, which adjusts for both fluoroscopy time and cumulative dose, told a converging story. Higher BMI again emerged as a significant predictor, as did longer fluoroscopy time and higher cumulative dose, each contributing independently to the total energy imparted across the irradiated field. Older age also predicted higher dose-area product, standing in contrast to its association with shorter fluoroscopy time. This divergence is instructive: in older patients, procedures may be quicker, but the dose delivered per unit time or per unit area can still be greater, perhaps reflecting anatomical changes or technical adjustments in beam settings. The lesson is that no single metric can stand in for the full radiation picture, and a procedure that looks efficient on one measure may still carry substantial dosimetric burden on another.

What do these findings mean in practice? The authors conclude that patient BMI is associated with overall radiation exposure, making body habitus a key variable to consider when planning and counseling patients before PTCD. Procedural complexity, reflected in the level of biliary obstruction, the choice of access route, and the severity of ductal dilatation, shows significant associations with fluoroscopy time and cumulative dose. In concrete terms, an interventional team facing an obese patient with severe ductal dilatation and a dual access plan can anticipate a higher radiation burden and take countermeasures: minimizing fluoroscopic runs, using last-image hold and other dose-saving features, collimating tightly to reduce the irradiated field area, and rotating staff positions to limit scatter exposure. Because scatter radiation to operators scales with the dose leaving the patient, every reduction in patient dose also translates into occupational protection.

The study’s design carries both strengths and limitations that readers should keep in view. Its use of three linked dosimetric metrics and a statistically careful modeling chain represents a methodological advance over single-outcome analyses, and the retrospective cohort of 100 patients from a single high-volume center provides internally consistent evidence. However, the authors themselves are explicit that these findings are exploratory and warrant validation in larger, prospective studies. A retrospective single-center design cannot capture every variable that influences radiation use, including operator experience, equipment generation, and institutional protocols, all of which may vary widely across hospitals. The effect sizes for some predictors, while statistically significant, are modest, and clinical decision-making should not rest on any single coefficient.

Even so, the study arrives at a moment when radiation safety in interventional radiology is under growing scrutiny. Procedures like PTCD are becoming more common as populations age and biliary disease burden rises, and cumulative occupational exposure among interventionalists is a persistent concern for the specialty. By identifying which patients and which procedural configurations predict the highest doses, the Beijing team has given clinicians a framework for pre-procedural radiation awareness, one that treats dosimetry not as an afterthought recorded at the end of a case but as a dimension of planning on par with access strategy and catheter selection. If larger prospective studies confirm these associations, the humble BMI measurement and the anatomical mapping performed before a needle ever touches the skin could become standard inputs into a personalized radiation risk estimate, turning routine clinical data into a shield for both patients and the physicians who treat them.

Subject of Research: Predictors of patient radiation exposure during percutaneous transhepatic cholangiodrainage

Article Title: Predictors of radiation exposure in percutaneous transhepatic cholangiodrainage: a multifactorial analysis integrating fluoroscopy time, cumulative dose, and dose-area product

Article References: Yang, B., & Jin, L. (2026). Predictors of radiation exposure in percutaneous transhepatic cholangiodrainage: a multifactorial analysis integrating fluoroscopy time, cumulative dose, and dose-area product. BMC Medical Imaging. https://doi.org/10.1186/s12880-026-02817-y

Image Credits: AI Generated

DOI: 10.1186/s12880-026-02817-y

Keywords: PTCD, radiation protection, fluoroscopy time, cumulative dose, dose-area product, interventional radiology, BMI, biliary drainage, radiation dosimetry, generalized linear models, Predictors, radiation

The patient’s background offered few warnings. He had undergone a laparoscopic cholecystectomy two years earlier, with surgical clips later confirmed to be in their expected position. He carried a diagnosis of poorly controlled type 2 diabetes, with a glycated hemoglobin of 9.7 percent, and took metformin twice daily. He drank no alcohol and had no known liver disease. Four days after his reassuring clinic visit, he underwent an elective ileocolonoscopy for chronic ileocaecal changes, which found three small ulcers in the terminal ileum and a rectal polyp that was removed. Crucially, the procedure involved no biliary instrumentation and no sphincterotomy, so it cannot plausibly explain what followed. On 16 February 2026, he presented acutely with two days of fever, progressive right upper quadrant and epigastric pain, nausea, and jaundice. He was tachycardic at 108 beats per minute but hemodynamically stable, with a blood pressure of 118 over 74 millimeters of mercury and a Glasgow Coma Scale score of 15. Charcot’s triad, the classic combination of fever, jaundice, and right upper quadrant pain, was complete.

The diagnostic machinery of TG18 then clicked into place. The guidelines define acute cholangitis across three domains: systemic inflammation, cholestasis, and imaging evidence of biliary obstruction or an etiology. This patient satisfied all three. He had fever with a white cell count of 14.2 thousand per microliter and C-reactive protein above 100 milligrams per liter. He had jaundice with a total bilirubin of 26 milligrams per deciliter, roughly 445 micromoles per liter, alongside markedly deranged liver enzymes. Triphasic computed tomography showed mild intrahepatic ductal dilatation and hyperattenuating dependent material in the distal common bile duct, three to four filling defects with the largest around four millimeters, suggesting sludge as the culprit. The severity grading, however, is where the case becomes genuinely instructive, because the arithmetic is a common source of clinical error.

TG18 Grade II, or moderate, cholangitis requires any two of five criteria: a white cell count above 12 or below 4 thousand per microliter, temperature of at least 39 degrees Celsius, age of 75 years or more, bilirubin of at least 5 milligrams per deciliter, and hypoalbuminemia defined as albumin below 0.7 times the local lower reference limit. This patient met two: the leukocytosis and the hyperbilirubinemia. He fell short on the other three, with a temperature of 38.9 degrees, an age of 42, and an albumin of 3.4 grams per deciliter. Two criteria are sufficient, so he was Grade II from the moment of diagnosis. Equally important is what he was not. No Grade III organ dysfunction was identified: his mean arterial pressure stayed at 89 millimeters of mercury or above without vasopressors, his creatinine of 1.1 milligrams per deciliter sat well below the 2.0 threshold, his INR of 1.3 was under the 1.5 cutoff, and his platelet count of 228 thousand per microliter exceeded the 100 thousand floor. The report candidly notes that two Grade III components, the respiratory criterion and oliguria, could not be formally assessed because no arterial blood gas was drawn and quantitative urine output was unavailable.

Grade II severity, established at presentation, is itself the indication for early biliary drainage alongside antimicrobial therapy. That is the guideline’s core message, and it is the point the case report hammers home: the drainage indication was created on day zero by the severity grade, not by any later failure to improve. Yet immediate endoscopic retrograde cholangiopancreatography, ERCP, was not feasible because of endoscopy-suite and anesthetic availability. What followed was a transparently documented interval rather than an unexamined one. The team started intravenous piperacillin-tazobactam at 4.5 grams every eight hours immediately after drawing blood cultures, in line with TG18 antimicrobial recommendations, and added fluid resuscitation, fasting, analgesia, and basal-bolus insulin. Severity was formally reassessed at 24 and 48 hours. By the 48-hour mark, fever and jaundice persisted, bilirubin remained at 24 milligrams per deciliter, ALT was still 837 units per liter, and CRP was 86 milligrams per liter. Those persistent abnormalities reinforced the urgency of drainage during the unavoidable delay, but the authors are careful to note they did not constitute a separate TG18 severity criterion and did not create the indication. Magnetic resonance cholangiopancreatography was deliberately skipped: the drainage indication was already established, and further imaging could not have altered it or expedited source control.

ERCP was performed on day three, the earliest available opportunity, and it delivered the definitive answer. The ampulla was edematous. Wire-guided cannulation of the common bile duct and sphincterotomy were carried out, and on aspiration, macroscopic purulent bile was identified at ductal entry, direct intra-procedural evidence of suppurative biliary infection. Fluoroscopy confirmed three to four distal filling defects, and three balloon sweeps extracted copious sludge and purulent material, with repeat imaging confirming clearance. The pancreatic duct was not cannulated and no stent was placed. The microbiology, however, is a lesson in itself. Blood cultures drawn before any antibiotic was given showed no growth at five days, and the bile culture, obtained after roughly 72 hours of piperacillin-tazobactam, was also sterile. The report explains why these two negative results mean different things: bacteraemia is documented in only a proportion of patients with acute cholangitis, so pre-antibiotic blood cultures are often negative, while antimicrobial exposure may have reduced the yield of the bile aspirate. Neither result excludes biliary infection, and neither implies that antibiotics reliably sterilize bile.

The biochemical trajectory raised another teaching point. Admission ALT hit 1020 units per liter, 46 times the patient’s documented baseline of 22 and 18.5 times the upper reference limit. Aminotransferase elevations of that magnitude often steer clinicians toward hepatocellular disease, but the report cites pooled data showing that marked elevations above 1000 units per liter occur in about 7.8 percent of acute choledocholithiasis cases. Two features favored biliary obstruction over ischemic hepatopathy: there was no hypotensive episode, and the enzymes declined in a progressive, stepwise fashion, with ALT falling 81 percent over seven days, rather than showing the massive but transient spike typical of hypoxic hepatitis. Pancreatic enzymes were also dramatically elevated, with amylase at 4830 units per liter, and the authors concede that concurrent mild biliary pancreatitis could not be excluded, since two of three Revised Atlanta criteria were satisfied, though the distinction did not change management. Viral hepatitis was ruled out serologically for hepatitis A, B, and C, with no new drugs or supplements identified, though hepatitis E testing was not performed.

Recovery after drainage was rapid. Fever resolved within 24 hours of ERCP, and by day five bilirubin had fallen to 14 milligrams per deciliter. He was discharged on day six, afebrile and eating normally. At three-month follow-up in mid-May 2026, every hepatobiliary, inflammatory, hematological, and renal parameter had normalized: ALT of 38 units per liter, bilirubin of 0.9 milligrams per deciliter, CRP below 5 milligrams per liter, and a white cell count of 6.8 thousand per microliter. His glycated hemoglobin improved from 9.7 to 8.3 percent, though it remained abnormal. He had suffered no recurrent biliary symptoms. The incidental terminal ileal biopsy was eventually reported as mild, nonspecific active chronic ileitis, with no causal role claimed in the biliary event.

The authors are unusually candid about the imperfections of the pathway they followed. Drainage occurred roughly 72 hours after admission, beyond the 48-hour window suggested by the American Society for Gastrointestinal Endoscopy for biliary decompression in cholangitis. They state plainly that day-three drainage is not a model Grade II pathway, that a fully resourced service would have drained sooner, and that the favorable outcome does not validate the delay. No antibiotics-first observation period, they warn, should be inferred from this recovery. They also list the case’s other limits: no organism was ever identified, the extracted sludge was not analyzed so a passed stone or microlithiasis cannot be excluded, the original ERCP images could not be retrieved from the institutional archive, and as a single case report it supports no causal inference or generalization.

What survives those caveats is the decision-making architecture itself. A documented normal hepatobiliary evaluation seven days earlier, and only mildly impressive CT findings on admission, could each have supplied false reassurance. The TG18 framework instead forced the team to grade severity prospectively, on objective criteria, and to let that grade, not the antecedent normality, drive management. Charcot’s triad, the report reminds readers, is a diagnostic construct that confers no severity grade; a patient can have a complete triad and still be Grade I, while another, like this one, reaches Grade II on leukocytosis and hyperbilirubinemia alone. In an era when cognitive bias quietly shapes clinical judgment, the case is a vivid argument for letting current severity, transparently graded and honestly reassessed, override every reassuring signal that came before.

Subject of Research: TG18-guided severity grading and timing of biliary drainage in ERCP-confirmed purulent cholangitis after a documented normal hepatobiliary evaluation

Article Title: When Current Severity Overrides Reassuring Findings: TG18‐Guided Biliary Drainage in ERCP‐Confirmed Purulent Cholangitis

Article References: Taha, H. M., & Taha, K. H. (2026). When Current Severity Overrides Reassuring Findings: TG18 ‐Guided Biliary Drainage in ERCP ‐Confirmed Purulent Cholangitis. Clinical Case Reports, 14(10), Article e73674. https://doi.org/10.1002/ccr3.73674

Image Credits: AI Generated

DOI: 10.1002/ccr3.73674

Keywords: acute cholangitis, Tokyo Guidelines 2018, biliary drainage, ERCP, biliary sludge, severity grading, Charcot's triad, hyperbilirubinemia, piperacillin-tazobactam, cholecystectomy, case report, clinical decision-making