📄 medicine

CUSUM Analysis of the Initial Learning Curve for Robot-Assisted Gynecologic Surgery at a German Tertiary Center

This retrospective study of 66 robot-assisted laparoscopic hysterectomies at a German tertiary center demonstrates that a single surgeon achieves procedural stabilization after approximately 36 cases, marked by a significant reduction in skin-to-closure time while maintaining comparable safety and perioperative outcomes.

Original authors: Morva Tahmasbi Rad, Dario Colacurci, Elias Bascharyar, Giuseppe Bifulco, Sven Becker, Ina Shehaj

Published 2026-08-07

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Morva Tahmasbi Rad, Dario Colacurci, Elias Bascharyar, Giuseppe Bifulco, Sven Becker, Ina Shehaj

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a world where surgeons don't just use their hands, but pilot a high-tech, three-dimensional video game controller that translates their movements into super-precise actions inside a patient's body. This is robotic-assisted surgery. Think of it like upgrading from a standard bicycle to a Formula 1 car: the steering is smoother, the view is crystal clear, and the machine filters out any tiny shakes in the driver's hands. In the field of gynecology, this technology is being used to perform delicate operations, like removing the uterus (a hysterectomy), with the goal of making recovery faster and less painful than traditional open surgery.

But here's the catch: even the best Formula 1 car needs a driver who knows how to handle it. When a surgeon first starts using a new robotic system, they are in the "learning curve." This is the period where they are figuring out the controls, the team is learning how to set up the machine quickly, and everyone is getting used to the new workflow. Just like learning to ride a bike, you wobble a bit at first, but eventually, you find your balance and start zooming. The big question for hospitals is: How many times does a surgeon need to do the procedure before they stop wobbling and start zooming? If they can figure this out, they can ensure patients get the best care right from the start, even when the surgery is tricky.

The Great Robot Hysterectomy Race

In a busy hospital in Frankfurt, Germany, a team of doctors decided to put this learning curve to the test. They wanted to see exactly how long it took for a single surgeon, working with a brand-new robotic team, to master the art of robot-assisted hysterectomy. They didn't just guess; they used a clever statistical tool called CUSUM analysis. Think of CUSUM like a "progress bar" on a video game. Instead of just looking at one level at a time, it tracks the total score over many levels to spot exactly when the player stops struggling and starts dominating.

The team watched the first 66 patients who underwent this surgery between January 2022 and September 2023. They measured three main things to see how the team was doing:

Skin-to-closure time: The total time from the first cut to the last stitch.
Console time: The actual time the surgeon spent sitting at the robot's controls.
Docking time: The time it took to roll the robot cart over to the patient and connect all the arms (like plugging in a gaming console).

The Big Discovery: The "36-Case" Magic Number
The progress bar told a very clear story. The team found a distinct turning point after the 36th case. Before this point (Cases 1–36), the team was in the "learning phase." After this point (Cases 37–66), they had hit "procedural consolidation," meaning they had found their rhythm.

The most obvious improvement was in the total surgery time. In the first 36 cases, the median time from skin incision to closing up was 108 minutes. Once they passed the 36-case mark, that time dropped significantly to 85.5 minutes. That's a saving of over 20 minutes per surgery! While the time the surgeon spent at the console and the time spent docking the robot also got faster, those improvements were a bit more subtle and didn't quite reach the level of statistical "significance" the team was looking for, though the trend was definitely positive.

It Wasn't Just About Speed
You might think that getting faster means cutting corners or taking risks, but the data says otherwise. The team was surprised to find that the patients in the "learning phase" were actually quite complex. Many had a history of previous abdominal surgeries, some had endometriosis (a painful condition where tissue grows where it shouldn't), and some had large uteruses or higher body mass indexes (BMI). In fact, 80.3% of the patients had had previous abdominal surgery, and 47% had endometriosis.

Despite these challenges, the "safety score" remained rock solid throughout the entire learning process.

Complications: The rate of problems after surgery (like infections) was low and didn't change between the early and late phases.
Conversions: Sometimes, a robot surgery has to be switched to a traditional open surgery or a standard laparoscopy. This happened in 10.6% of cases (7 out of 66). Interestingly, most of these weren't because the robot failed; they happened because the doctors found a large uterus that might be cancerous and needed to be removed whole without breaking it apart. Only one conversion was due to the robot struggling with severe scar tissue (adhesions).
Pain and Stay: Patients felt about the same amount of pain and stayed in the hospital for the same amount of time, regardless of whether they were in the first half or the second half of the study.

The Team Effect
One of the coolest findings was that the learning curve wasn't just about the surgeon's hands. The "docking time" (setting up the robot) improved as the whole team—nurses, assistants, and the surgeon—got better at working together. The data showed that the setup time peaked around case 28 and then started to drop, suggesting that the team needed to learn how to coordinate just as much as the surgeon needed to learn the controls.

What This Means
The paper concludes that robotic surgery can be introduced safely, even for complex cases, as long as there is a structured plan. The surgeon and the team didn't need to wait until they were "experts" to start helping patients; they just needed to get through that first 36 cases to find their groove. After that, the surgeries became noticeably faster without sacrificing safety.

So, the next time you hear about a new medical technology, remember the story of the 66 patients and the magic number 36. It's a reminder that with practice, teamwork, and a little bit of patience, even the most high-tech tools can become second nature, leading to better, faster care for everyone.

Technical Summary: CUSUM Analysis of the Initial Learning Curve for Robot-Assisted Gynecologic Surgery

Problem Statement
The adoption of robotic-assisted surgery (RAS) in gynecology requires a transition from conventional minimally invasive techniques to a new technological platform. While RAS offers technical advantages such as 3D visualization and tremor filtration, its superiority over traditional laparoscopy remains debated, particularly regarding operative time and institutional costs. A critical factor in successful implementation is the "learning curve"—the period required for a surgical team to acquire technical competence and optimize workflow. There is a need for objective evaluation of this process to ensure patient safety and procedural efficiency, especially in complex cases involving obesity, prior abdominal surgery, or large uterine volumes. This study addresses the gap in understanding the specific learning trajectory of a single surgeon and their operating room team during the initial implementation of robotic gynecologic surgery at a German tertiary university center.

Methodology
This retrospective observational study analyzed the first 66 consecutive patients who underwent robot-assisted laparoscopic hysterectomy using the da Vinci X Surgical System between January 2022 and September 2023 at the University Hospital Frankfurt. All procedures were performed by a single primary console surgeon (M.T.R.) and the same operating room team, neither of whom had prior independent robotic experience.

Data Collection: Variables included demographic data, surgical complexity indicators (BMI, uterine volume, adhesion scores, endometriosis), and operative metrics (skin-to-closure time, console time, docking time, positioning time).
Statistical Analysis: The primary analytical tool was Cumulative Sum (CUSUM) analysis. This method monitored operative performance over time to identify inflection points where procedural stabilization occurred.
- The formula used was: $CUSUM(S_n) = \sum(X_i - X_0)$ , where $X_i$ is the operative time for a specific case and $X_0$ is the mean operative time of the entire series.
- Based on the CUSUM curve for skin-to-closure time, the cohort was divided into two phases: Phase 1 (cases 1–36) representing the initial learning phase, and Phase 2 (cases 37–66) representing procedural consolidation.
Comparative Analysis: Outcomes between the two phases were compared using Mann–Whitney U tests for continuous variables and Chi-square or Fisher's exact tests for categorical variables.

Key Contributions and Results
The study provides a quantified learning curve for robotic hysterectomy in a high-complexity setting, demonstrating that procedural proficiency can be achieved safely within a structured training framework.

Learning Curve Identification: CUSUM analysis identified a distinct transition point at case 36 for skin-to-closure time. This marks the completion of the initial learning phase and the onset of procedural stabilization.
Operative Efficiency:
- Skin-to-closure time: Decreased significantly from a median of 108 minutes in Phase 1 to 85.5 minutes in Phase 2 ( $p = 0.029$ ).
- Console time: Showed a favorable trend, decreasing from 66.5 to 62 minutes, though not statistically significant ( $p = 0.074$ ).
- Docking time: Declined from 24 to 18.5 minutes, with the CUSUM curve peaking at case 28 before trending downward, indicating team coordination maturation.
- Positioning time: Significantly increased in Phase 2 ( $p = 0.008$ ), attributed to the inclusion of more complex cases (higher BMI, larger uterine volume) requiring more extensive preparation.
Safety and Complexity:
- The cohort was complex: 80.3% had prior abdominal surgery, 47% had endometriosis, and 41% had moderate-to-severe adhesions.
- Conversion Rate: The overall conversion rate was 10.6% (7/66), with no significant difference between phases. Notably, most conversions were not due to technical failure but were safety decisions to avoid morcellation in cases with suspected malignancy or large uterine size.
- Complications: The overall complication rate was 7.6% (wound infection, urinary tract infection, pneumothorax), with no significant difference between phases.
- Outcomes: Postoperative pain scores, length of hospital stay, and blood loss remained stable and comparable across both phases.

Significance and Claims
The paper claims that robot-assisted laparoscopic hysterectomy demonstrates a well-defined learning curve, with procedural stabilization achieved after approximately 36 consecutive cases.

Feasibility in Complex Cases: The study asserts that the successful and safe implementation of robotic surgery is feasible during the initial adoption phase, even for complex cases (high BMI, large uterine volume, adhesions), provided there is structured training and standardized workflows.
Team Dependency: The findings highlight that robotic implementation is not solely surgeon-dependent but also team-dependent. The maturation of docking and positioning times suggests that the entire operating room team (assistants, nurses, anesthesiologists) undergoes a learning curve.
Safety First: The authors emphasize that maintaining safety during the learning phase is paramount. The study supports the concept that robotic surgery can be safely introduced for technically demanding indications if case selection is careful and the team follows a certified training pathway.
Methodological Value: The use of CUSUM analysis is presented as a robust method for objectively visualizing performance transitions and identifying the specific number of cases required to move from the acquisition phase to procedural consolidation.

The authors conclude that while operative time improves significantly after the learning curve, the primary goal of implementation should be the safe extension of minimally invasive approaches to complex patient populations rather than simply replacing laparoscopy in routine cases.

The Great Robot Hysterectomy Race

Technical Summary: CUSUM Analysis of the Initial Learning Curve for Robot-Assisted Gynecologic Surgery

More like this