Cumulative summation (CUSUM) charts in the monitoring of hypospadias outcomes: a tool for quality improvement initiative

Cumulative sum learning curve analysis of tubularized incised plate repair for hypospadias: a study of a single surgeon with a single surgical procedure

Purpose

To ascertain the quantity of instances by which a single surgeon achieves competency and proficiency in using tubularized incised plate (TIP) technique for the repair of distal and mid-shaft hypospadias using the cumulative sum (CUSUM) analysis.

Methods

We retrospectively evaluated patients with distal and mid-shaft hypospadias who were treated by a single surgeon between 2015 and 2021, using a single primary TIP technique with a de-epithelialized Byars flap. Data including type of hypospadias, age at surgery, curvature, operation time (OT), length of the reconstructed urethra, and postoperative outcomes were collected and assessed. CUSUM was used to assess the trends in OT and complication rate (CR) in order to generate the learning curve. The evolution of OT and CR can be divided into three phases: learning, competence, and proficiency.

Results

CUSUM identified three phases in the learning curves of all TIP repairs. The median OT decreased from 135 min [interquartile range (IQR) = 125-155] to 92 min (IQR = 80-100) (P < 0.001), CR decreased from 28 (28%) to 8 (5.3%) (P < 0.001), and reoperations decreased from 15 (15.2%) to 4 (2.6%) (P < 0.001). According to the CUSUM learning curve, technical competency plateaued after the 99th case, and both OT and CR entered a significantly declining proficiency phase after the 231st case. Further, when the neourethral length exceeded the total average, total complications, urethrocutaneous fistula, and reoperations increased (P = 0.013, P = 0.006, and P = 0.028, respectively).

Conclusions

Our study suggests that surgeons performing TIP repair may reach technical competency and achieve proficiency after operating on 99,231 cases, respectively. Moreover, the longer the neourethral length, the higher is the CR.

Significance of data analysis in the quality control of prenatal screening for Down syndrome

Dual detection of α-fetoprotein (AFP) and free β-human chorionic gonadotropin (β-HCG) is a common screening method for Down syndrome in the second trimester and its efficacy is assessed by false-positive rate (FPR). The present study aimed to investigate the effects of the bias in median multiple of the median (mMoM) values of AFP and free β-HCG on FPR. The bias in mMoM values of AFP and free β-HCG and the bias in mMoM values under different gestational ages and weight groups were analyzed. Median equations were adjusted, and medians in LifeCycle software were replaced by local medians. Following two adjustments of the median equations, all indices including FPR, mMoM values of markers and mMoM values under different gestational ages and weight groups generally reached an ideal state. In conclusion, abnormal bias in mMoM values may prompt aberrant application of median equations, and regular monitoring of these indicators may be important for quality control in prenatal screening.

A Bayesian CUSUM plot: Diagnosing quality of treatment

OBJECTIVES

To present a CUSUM plot based on Bayesian diagnostic reasoning displaying evidence in favour of "healthy" rather than "sick" quality of treatment (QOT), and to demonstrate a technique using Kaplan-Meier survival curves permitting application to case series with ongoing follow-up.

METHODS

For a case series with known final outcomes: Consider each case a diagnostic test of good versus poor QOT (expected vs. increased failure rates), determine the likelihood ratio (LR) of the observed outcome, convert LR to weight taking log to base 2, and add up weights sequentially in a plot showing how many times odds in favour of good QOT have been doubled. For a series with observed survival times and an expected survival curve: Divide the curve into time intervals, determine "healthy" and specify "sick" risks of failure in each interval, construct a "sick" survival curve, determine the LR of survival or failure at the given observation times, convert to weights, and add up.

RESULTS

The Bayesian plot was applied retrospectively to 39 children with acute lymphoblastic leukaemia with completed follow-up, using Nordic collaborative results as reference, showing equal odds between good and poor QOT. In the ongoing treatment trial, with 22 of 37 children still at risk for event, QOT has been monitored with average survival curves as reference, odds so far favoring good QOT 2:1.

CONCLUSION

QOT in small patient series can be assessed with a Bayesian CUSUM plot, retrospectively when all treatment outcomes are known, but also in ongoing series with unfinished follow-up.

Learning curve evaluation using cumulative summation analysis-a clinical example of pediatric robot-assisted laparoscopic pyeloplasty

BACKGROUND

The cumulative summation (CUSUM) method for learning curve analysis remains under-utilized in the surgical literature in general, and is described in only a small number of publications within the field of pediatric surgery. This study introduces the CUSUM analysis technique and applies it to evaluate the learning curve for pediatric robot-assisted laparoscopic pyeloplasty (RP).

METHODS

Clinical data were prospectively recorded for consecutive pediatric RP cases performed by a single-surgeon. CUSUM charts and tests were generated for set-up time, docking time, console time, operating time, total operating room time, and postoperative complications. Conversions and avoidable operating room delay were separately evaluated with respect to case experience. Comparisons between case experience and time-based outcomes were assessed using the Student's t-test and ANOVA for bi-phasic and multi-phasic learning curves respectively. Comparison between case experience and complication frequency was assessed using the Kruskal-Wallis test.

RESULTS

A total of 90 RP cases were evaluated. The learning curve transitioned beyond the learning phase at cases 10, 15, 42, 57, and 58 for set-up time, docking time, console time, operating time, and total operating room time respectively. All comparisons of mean operating times between the learning phase and subsequent phases were statistically significant (P=<0.001-0.01). No significant difference was observed between case experience and frequency of post-operative complications (P=0.125), although the CUSUM chart demonstrated a directional change in slope for the last 12 cases in which there were high proportions of re-do cases and patients <6 months of age.

CONCLUSIONS

The CUSUM method has a valuable role for learning curve evaluation and outcome quality monitoring. In applying this statistical technique to the largest reported single surgeon series of pediatric RP, we demonstrate numerous distinctly shaped learning curves and well-defined learning phase transition points.