Can Computed Tomography Findings for Kidney, Ureter and Bladder Correlate with Medical Comorbidity in Renal Colic Patients?

Abstract

Background/Objectives: Sarcopenia is a progressive skeletal muscle disorder linked to adverse outcomes. Computed Tomography (CT) can quantify skeletal muscle, while the Charlson Comorbidity Index (CCI) predicts mortality by categorising comorbidities. This study examined whether Computed Tomography of the Kidneys, Ureters, and Bladder (CT-KUB)-derived skeletal muscle measurements correlate with CCI scores in hospitalised patients. Methods: This retrospective study included all patients admitted with renal colic to the Urology Department, Blacktown Hospital and underwent cystoscopy between June 2022 and June 2025. Data were obtained from electronic medical records. CCI scores, incorporating age and comorbidities, generated 10-year survival estimates. CT-KUB scans were reviewed for psoas muscle perimeter, area, height, width and Hounsfield unit at the aortic bifurcation. Skeletal Muscle Index (SMI) was calculated as skeletal muscle area (SMA)/height². Associations between CCI, psoas muscle metrics and outcomes (length of stay, Intensive Care Unit (ICU) admission, Emergency Department (ED) re-presentation) were assessed using Pearson’s correlations and between-group comparisons. Results: A total of 397 patients were analysed. Median Length of Stay (LOS) was 1 day (mean = 1.92, SD = 1.88). ICU admission occurred in 2.3% of patients, and 18.6% re-presented to ED within 30 days. Both CCI survival percentage and psoas muscle metrics (including SMI) were significantly associated with LOS. Lower SMA, Hounsfield unit (HU), length and perimeter were linked to higher ICU admission risk. Neither CCI nor muscle measures predicted ED re-presentation. Conclusions: CCI and CT-derived muscle metrics were independently associated with outcomes such as LOS and ICU admission. Combining these measures may improve risk stratification, warranting further prospective evaluation.

1. Introduction

Body composition, particularly the quality and quantity of skeletal muscle, is increasingly important when assessing patient health outcomes. Sarcopenia is a progressive and generalised skeletal muscle disorder associated with increased likelihood of adverse outcomes, including falls, fractures, physical disability and mortality [1]. The European Working Group on Sarcopenia in Older People 2 (EWGSOP2) defines sarcopenia as a muscle disease characterised by low muscle strength, low muscle quantity or quality, and poor physical performance [1]. Sarcopenia has also been shown to be a predictor of poor outcomes in oncology populations and other patient groups [2]. Skeletal muscle quantity can be measured as a skeletal muscle index (SMI) which is derived from total skeletal muscle area (SMA) divided by the square of the patient’s height [3]. SMA can be obtained from cross-sectional computed tomography (CT) imaging, most commonly at the T11–T12 [3,4,5,6] or L3 vertebral levels [7,8,9].

CT imaging is a non-invasive procedure that can accurately assess body composition measurements, such as SMA, SMI and radiodensity, which have been linked to increased morbidity and mortality in various patient populations [7,10,11,12]. Computed Tomography of the Kidneys, Ureters, and Bladder (CT-KUB) scans are the current standard approach for patients presenting with acute renal colic symptoms to the emergency department. This offers an underused opportunity to assess these metrics using readily available and pre-existing software [4,5,6]. The choice to use CT-KUB scans in this study was based on the following three reasons:

• Availability and feasibility: Blacktown Hospital’s electronic medical record (eMR) already contains a large database of CT-KUB scans, which are routinely performed for suspected renal colic [6].
• Representative sampling: CT-KUB scans are performed without restrictions on age, sex, or ethnicity, ensuring a representative sample of the hospital’s patient population.
• Standardisation of measurement: CT-KUB scans consistently include the T10–T12 vertebral levels, enabling uniform skeletal muscle measurements across patients [4,5].

The Charlson Comorbidity Index (CCI) is a validated tool for predicting mortality by categorising comorbid conditions using administrative data [13]. While both CCI and sarcopenia independently predict adverse outcomes, combining CT-derived body composition metrics with CCI scores has the potential to enhance predictive accuracy by accounting for physical resilience, which is critical for patient outcomes [14,15]. This study aims to evaluate whether CT-KUB-derived skeletal muscle measurements correlate with CCI scores in hospitalised patients, supporting a low-cost and comprehensive risk assessment model.

2. Materials and Methods

2.1. Study Design and Setting

This retrospective analysis was conducted on all patients admitted to the Department of Urology, Blacktown Hospital between June 2022 and June 2025 with renal colic and who underwent a cystoscopy as the primary intervention. By selecting this subset of patients (i.e., patients presenting with symptoms of renal colic to the Emergency Department), the analysis focused on a population where the only differences among individuals was demographics and pre-existing comorbidities. Ethics approval was obtained on the 12 December 2025 through the Western Sydney Local Health District Human Research Ethics Committee [ETH01502].

2.2. Study Population

A total of 1477 Emergency Department (ED) admissions via Urology were screened (Figure 1). Of these, 516 were excluded due to non-renal colic presentations (e.g., abscesses, torsion, orchitis). The remaining 916 patients had clinically suspicious renal colic based on ED assessment.

Among these, 578 patients had CT-KUB imaging available and accessible in Sectra at the time of admission. Cross-referencing against the 1346 surgical urology cases during the study period yielded 453 patients found on both lists.

After further exclusion of patients with non-renal colic presentation (n = 33), post-surgical complications (n = 18), inter-hospital transfers (n = 2), poor-quality scans (n = 2) or discharge prior to procedure (n = 1), a total of 397 patients were included in the final analysis (Figure 1). Within this group, 187 were also found to have height available, enabling calculation of SMI.

Inclusion criteria:

• All adult patients who presented to the Blacktown Emergency Department with renal colic AND were admitted to the Urology Department AND underwent a cystoscopy as the primary intervention.

Exclusion criteria:

• Poor-quality or incomplete scans (i.e., not including the level of aorta bifurcation).
• Insufficient clinical and demographic data to calculate the CCI.

Although all patients met inclusion criteria of suspected renal colic requiring admission and cystoscopy, variability in disease severity (including stone burden, degree of obstruction, presence of infection and clinical acuity) was not formally stratified. This reflects real-world emergency urology practice but may introduce unmeasured confounding.

2.3. Data Collection

Data were obtained from the eMR database. The dataset was provided by the Western Sydney Local Health District (WSLHD) clinical analytics team, who ran reports on urology admission data according to the inclusion criteria.

The following variables were recorded:

• Demographics: age, sex.
• Anthropometrics: height (for calculation of skeletal muscle index).
• Comorbidities.
• Imaging metrics: perimeter and area (minimum 8 points circumscribing the psoas on a single axial slice), height, width and Hounsfield unit of psoas muscle at the level of aorta bifurcation.
• Outcomes: length of stay, Intensive Care Unit (ICU) admission, 30-day readmission.

The CCI calculator [16] was used, incorporating age and comorbidities including myocardial infarction, congestive heart failure, peripheral vascular disease, cerebrovascular accident or transient ischemic attack, dementia, chronic pulmonary disease, connective tissue disease, peptic ulcer disease, liver disease, diabetes mellitus, hemiplegia, moderate to severe chronic kidney disease, solid tumour, leukaemia, lymphoma and Acquired Immunodeficiency Syndrome (AIDS). An estimated 10-year survival score was generated.

2.4. Imaging Analysis

CT-KUB scans were reviewed and measured for perimeter, area, height, width and the Hounsfield unit of the psoas muscle at the level of aorta bifurcation. SMI was calculated as: $\mathrm{S}\mathrm{M}\mathrm{I}={\displaystyle \frac{\mathrm{S}\mathrm{M}\mathrm{A}}{{\mathrm{Height}}^2}}$. SMA was obtained from imaging using minimum 8-point circumscription of the psoas muscle. For each measurement, both the right and left psoas muscles were collected. For area, perimeter, length, width and Hounsfield unit (HU), these values were averaged to yield a single measurement per patient. In the calculation of SMI, the sum of the right and left SMA was used. This was done to reduce variability and was consistent with prior CT-based sarcopenia studies where the total psoas area or mean attenuation from both sides was utilised to improve reliability [10,17,18].

2.5. Statistical Analysis

All statistical analysis was performed using IBM SPSS Statistics (Version 30, Armonk, NY, USA) [19]. To explore the associations between CCI (expressed as ten-year survival percentage), psoas muscle metrics (SMA, HU, length, width, perimeter), SMI and clinical outcomes including length of stay (LOS), ICU admission, ED re-presentation and age, Pearson’s correlation coefficients (r) were used. Between-group comparisons (e.g., ICU admission, ED re-presentation, sex) were performed using independent samples t-tests for normally distributed variables, and effect sizes were calculated where appropriate. Statistical significance was set at p < 0.05. Additionally, descriptive statistics were reported as means with standard deviations (SDs) or medians with interquartile ranges (IQRs) for continuous variables and as frequencies with percentages for categorical variables.

Multivariable analysis was performed to assess the independent association between comorbidity burden and CT-derived muscle metrics with clinical outcomes. Linear regression was used for LOS, while binary logistic regression was used for ICU admission and ED re-presentation. Models were adjusted for age and sex. A secondary analysis was performed in the subset of patients with available height data using skeletal muscle index.

2.6. Data Handling and Confidentiality

Ethics approval was obtained through the Western Sydney Local Health District Human Research Ethics Committee [ETH01502]. At the point of data entry, each patient was assigned a unique, random 3-alphanumeric code. The working dataset did not contain identifiers such as medical record numbers, names, or dates of birth. All data was stored in a password-protected Microsoft Excel sheet on a password-protected NSW Health computer, with access restricted to authorised personnel only.

3. Results

3.1. Patient Characteristics

A total of 397 patients were included in the study, of whom 60.7% were male and 39.3% were female. The mean age at admission was 50.37 years (SD = 14.57). The median LOS was 1 day (mean = 1.92, SD = 1.88). ICU admission occurred in 2.3% of patients, and 18.6% experienced an ED re-presentation within 30 days. The univariate associations between CCI, psoas muscle metrics, skeletal muscle index, and clinical outcomes are summarised in

Table 1. Correlations of Charlson Comorbidity Index (10-year survival %), psoas muscle metrics and Skeletal Muscle Index with clinical outcomes.

Predictor	LOS r (p)	ICU Admission	ED Re-Presentation	Age r (p)
CCI (10-yr survival %)	−0.303 (<0.001)	ns (0.393)	ns (0.854)	−0.676 (<0.001)
SMA (cm2)	−0.293 (<0.001)	99.3 vs. 130.8 (0.023)	ns (0.402)	−0.161 (0.001)
HU	−0.270 (<0.001)	37.5 vs. 43.6 (0.029)	ns (0.628)	−0.428 (<0.001)
Length (cm)	−0.288 (<0.001)	3.6 vs. 4.2 (0.009)	ns (0.980)	ns (0.066)
Width (cm)	−0.281 (<0.001)	ns (0.070)	ns (0.272)	−0.205 (<0.001)
Perimeter (cm)	−0.283 (<0.001)	12.2 vs. 14.0 (0.007)	ns (0.731)	ns (0.069)
SMI (subset, n = 187)	−0.347 (<0.001)	ns (0.706)	ns (0.975)	not reported

CCI: Charlson Comorbidity Index; SMA: skeletal muscle area; HU: Hounsfield unit; ns: not significant; SMI: skeletal muscle index; ICU: intensive care unit; ED: emergency department; LOS: length of stay.

.

3.2. Association Between Charlson Comorbidity Index and Clinical Outcomes

Higher CCI-predicted ten-year survival percentage was significantly correlated with shorter LOS (r = −0.303, p < 0.001). CCI survival percentage also demonstrated a strong negative correlation with age (r = −0.676, p < 0.001). No significant correlations were observed between CCI and ICU admission (r = −0.043, p = 0.393) or ED re-presentation within 30 days (r = 0.009, p = 0.854).

3.3. Association Between Psoas Muscle Metrics and Clinical Outcomes

All psoas muscle metrics were negatively correlated with LOS, including cross-sectional area (r = −0.293, p < 0.001), mean HU (r = −0.270, p < 0.001), length (r = −0.288, p < 0.001), width (r = −0.281, p < 0.001) and perimeter (r = −0.283, p < 0.001).

Psoas HU correlated positively with CCI mortality percentage (r = 0.345, p < 0.001) and negatively with age (r = −0.428, p < 0.001). Psoas width was also negatively correlated with age (r = −0.205, p < 0.001). No psoas muscle metric was significantly associated with ED re-presentation.

3.4. Combined Predictive Value of Comorbidity and Muscle Metrics

When considered together, comorbidity burden and muscle characteristics demonstrated predictive value for hospital outcomes. ICU admission was associated with adverse muscle profiles, including smaller cross-sectional area (mean 99.3 vs. 130.8 cm², p = 0.023), lower HU (mean 37.5 vs. 43.6, p = 0.029), shorter psoas length (mean 3.6 vs. 4.2 cm, p = 0.009) and reduced perimeter (mean 12.2 vs. 14.0 cm, p = 0.007). Patients admitted to the ICU also had a significantly longer LOS compared with non-ICU patients (mean 6.6 vs. 1.8 days, p < 0.001).

3.5. Skeletal Muscle Index

In the subset of patients with available SMI measurements, lower SMI was significantly associated with longer LOS (r = −0.347, p < 0.001). No significant associations were observed between SMI and ICU admission (r = −0.028, p = 0.706) or ED re-presentation (r = 0.002, p = 0.975).

3.6. Multivariable Analysis

Multivariable analyses were performed, adjusting for age and sex, with representative muscle variables selected to avoid redundancy between related measurements (

Table 2. Multivariable regression analysis of predictors of clinical outcomes. CI—Confidence Interval, OR—Odds Ratio.

Length of Stay (Linear Regression)
Variable	Effect Estimate (β)	95% CI	p-Value
CCI (10 year survival %)	−0.019	−0.029 to −0.009	<0.001
HU	−0.026	−0.049 to −0.002	0.033
SMA (cm2)	−0.006	−0.012 to −0.000	0.056
Age	0.001	−0.008 to −0.010	0.809
Sex (male)	−0.374	−0.768 to 0.021	0.061
30-Day ED Re-Presentation (Logistic Regression)
Variable	OR	95% CI	p-Value
CCI	1.00	0.98 to 1.02	0.987
HU	0.99	0.96 to 1.02	0.437
SMA (cm2)	1.01	0.99 to 1.02	0.219
Age	1.00	0.98 to 1.02	0.874
Sex (Male)	0.86	0.57 to 1.30	0.480
SMI Subset (n = 187)—Length of Stay
Variable	Effect Estimate (β)	95% CI	p-Value
CCI	−0.020	−0.038 to −0.003	0.020
SMI	−0.022	−0.037 to −0.006	0.008
HU	−0.041	−0.079 to −0.003	0.036

CCI: Charlson Comorbidity Index; SMA: skeletal muscle area; HU: Hounsfield unit; SMI: skeletal muscle index; ED: emergency department.

). On linear regression, higher CCI survival percentage (β = −0.019, p < 0.001) and higher psoas attenuation (as indicated by HU) (β = −0.026, p = 0.033) remained independently associated with shorter length of stay, while skeletal muscle area demonstrated a borderline association (β = −0.006, p = 0.056). Age and sex were not independently associated with LOS. On logistic regression, no variable independently predicted 30-day ED re-presentation. ICU admission was not included in the multivariable analysis due to the low number of events (n = 9), which limited model stability. In the subset of patients with available SMI, both SMI (β = −0.022, p = 0.008) and HU (β = −0.041, p = 0.036) remained independently associated with length of stay, alongside CCI (β = −0.020, p = 0.020).

4. Discussion

The prior literature supports an association between increased comorbidity burden and prolonged hospitalisation [13,14]. Further, sarcopenia has been identified as a predictor of poor postoperative recovery and increased ICU utilisation [20,21]. Most prior research, however, has been conducted in surgical, oncology, or critical-care cohorts [2,15,20,21,22,23], so the extent to which these findings hold in a general hospital population has not been fully established. This study evaluated the association between comorbidity burden, CT-derived muscle metrics, and clinical outcomes in patients admitted with renal colic. In this cohort of 397 patients, both CCI (expressed as the ten-year survival percentage) and psoas muscle metrics were associated with LOS on univariate analysis. Several muscle metrics were also associated with ICU admission, whereas neither CCI nor muscle metrics were associated with 30-day ED re-presentation

Multivariable analysis demonstrated that both comorbidity burden and CT-derived muscle metrics remained independently associated with LOS after adjustment for age and sex. This suggests that these measures capture distinct aspects of patient health, with comorbidity reflecting chronic disease burden and muscle metrics reflecting physiological reserve. Notably, psoas muscle attenuation remained independently associated with LOS, whereas skeletal muscle area demonstrated only a borderline association. This finding supports the concept that muscle quality, rather than quantity alone, may better reflect functional status and physiological resilience in acute illness.

Within the subset of 187 patients in whom SMI could be calculated, lower SMI correlated with longer LOS, consistent with findings from oncology and critical-care populations [2,22]. However, SMI was not associated with ICU admission or ED re-presentation, likely reflecting the reduced sample size and low number of ICU events. ED re-presentation is also influenced by additional factors, including access to care, procedural complications and broader patient-level variables [24,25,26]. Although several muscle metrics were associated with ICU admission on univariate analysis, no independent predictors were identified on multivariable modelling. This is likely attributable to the low number of ICU events in this cohort, which limited statistical power, and these findings should therefore be interpreted with caution. Similarly, no independent predictors of 30-day ED re-presentation were identified, suggesting that this outcome is likely influenced by factors not captured in this dataset, including access to care, procedural factors, and social determinants.

These findings reinforce the prognostic value of muscle mass and quality for acute outcomes, and they align with prior work linking sarcopenia to poor recovery and greater ICU use [20,21]. In contrast, the CCI survival percentage, being a long-term predictor, was not predictive of ICU admission or 30-day ED re-presentation. This suggests that although comorbidity burden influences recovery duration, it may not adequately capture the acute physiological reserve required for critical illness or readmission risk. Taken together, these results indicate that muscle metrics and comorbidity indices may be able to provide distinct but complementary insights into both immediate and longer-term prognoses for patients.

Our analysis further demonstrated that meaningful correlations can be obtained without specialised third-party software (e.g., SliceOmatic, ImageJ, OsiriX) which requires training, eMR integration and, often, licensing [2,15]. For our study, psoas muscle parameters were derived directly from CT-KUB images using a standard viewer. Despite this, the associations observed were comparable to prior studies using dedicated platforms [15], demonstrating that the measurements can be performed reliably, without additional cost and can even be extracted by a supervised medical student, as was done in our cohort. It would be beneficial for future research to compare manual, non-proprietary methods with specialised software to assess prognostic value beyond what can be achieved with routine image review.

This study has several limitations. The retrospective, single-centre design may limit generalisability. Although multivariable analysis was performed, residual confounding from unmeasured factors such as stone burden, degree of obstruction, and presence of infection remains possible. Additionally, SMI could only be calculated in a subset of patients due to incomplete anthropometric data, which may reduce statistical power. Nevertheless, the consistency of findings in this subgroup support the robustness of the primary analysis. The low number of ICU events limited the ability to detect independent predictors for critical care outcomes.

The study cohort reflects a pragmatic emergency urology population rather than a strictly homogeneous disease entity. Although all patients presented with suspected renal colic and underwent insertion of ureteric stent, variability in stone characteristics, degree of obstruction, and presence of infection may influence outcomes such as LOS and ICU admission. These factors were not captured in the current dataset and may act as confounders. Future studies incorporating radiological stone parameters and markers of sepsis are required to better delineate these effects.

Incorporating CT-derived muscle metrics alongside comorbidity indices may improve early identification of patients at risk of prolonged hospitalisation. Given that CT-KUB imaging is routinely performed in this population, opportunistic assessment of muscle attenuation represents a practical and cost-neutral approach to enhancing clinical risk stratification.

5. Conclusions

The findings of this study indicated that comorbidity burden was associated with longer hospital stay, and psoas muscle metrics and SMI provided comparable or stronger predictive value for length of stay. Several psoas measures also showed potential for predicting ICU admission. Neither CCI nor muscle metrics predicted ED re-presentation. These findings support integrating image-based muscle metrics into risk stratification alongside comorbidity indices. Further larger, multicentre and prospective studies incorporating disease specific variables, longer follow-up periods and hospital economic evaluation are required to validate these findings.