Abstract
One objective of ophthalmological departments is the optimization of patient treatment services. A strategy for optimization is the identification of individual potential for advanced training of surgeons based on their daily working results. The objective of this feasibility study was the presentation and evaluation of a strategy for the computation of surgeon–individual treatment profiles (SiTPs). We observed experienced surgeons during their standard daily performance of cataract procedures in the Ophthalmological Department of the University Medical Center Leipzig, Germany. One hundred five cases of cataract procedures were measured as Surgical Process Models (SPMs) with a detailed-to-the-second resolution. The procedures were performed by three different surgeons during their daily work. Subsequently, SiTPs were computed and analyzed from the SPMs as statistical ‘mean’ treatment strategies for each of the surgeons. The feasibility study demonstrated that it is possible to identify differences in surgeon–individual treatment profiles beyond the resolution of cut–suture times. Surgeon–individual workflows, activity frequencies and average performance durations of surgical activities during cataract procedures were analyzed. Highly significant (p < 0.001) workflow differences were found between the treatment profiles of the three surgeons. Conclusively, the generation of SiTPs is a convenient strategy to identify surgeon–individual training potentials in cataract surgery. Concrete recommendations for further education can be derived from the profiles.
Keywords: Education; Surgical procedures, operative; Ophthalmology; Process assessment (health care); Workflow; Surgical process model
Introduction
The performance of surgical procedures is a complex interaction of manual skills and the experience that surgeons gather in the course of their professional life [1, 2]. The mediation of this accumulated knowledge to junior surgeons usually ensues by means of oral information, a one to one observation of the senior surgeons by the residents with subsequent practice, or other ‘knowledge stores’, such as clinical guidelines [3, 4] or dedicated literature and videos for surgical education.
It is the objective of each clinic to ensure not only an optimal, but also a homogenous, treatment service for every patient. However, due to patient-specific characteristics, different techniques favored by different seniors, to varying preferences and the availability of technical resources to support the surgical task, a variation of surgical procedure courses results. To achieve a homogenous success of treatment, it is an option to minimize the variability in treatment service caused by different surgical preferences. With relation to the individual skill and experience of the surgeons, it is necessary to identify the individual potential for improvement of their surgical abilities and a subsequent targeted support with the help of advanced surgical training.
The goal of the presented feasibility study was the presentation of a method and its evaluation for the computation of the individual promotion potential for surgeons in their working life by advanced training. In the course of the study, the surgical processes of 105 cataract procedures that were performed by three different surgeons were measured and analyzed. The promotion potential for each of the three surgeons was investigated and identified. The objective of the feasibility study was to answer the questions: (1) ‘How can surgeon–individual treatment service can be assessed?’, (2) ‘How can a statistically averaged surgeon–individual treatment profile (SiTP) be identified?’ and (3) ‘How can these SiTPs be used to derive advanced training strategies for the surgeon?’
The major focus of publications on the education and the training of cataract surgery was on the design and the evaluation of training programs for residents, such as the training of residents by a virtual mentor system [5], the training of complication management [6], the impact of the residents’ curriculum design on complication rates [7] or the use of models to teach residents surgical work steps in eye surgery [8, 9] .Furthermore, a number of works have focused on the objective assessment of surgical skills with advanced methods [10–12] or on data acquisition strategies in ophthalmology training [13, 14].
However, none of these works have used explicitly measured surgical processes to assess and support the education and training and to achieve detailed improvement results. The explicit measuring and modeling of the surgical processes is a relatively new research topic. So far, the application scenarios for surgical process models have mainly been the optimization of surgical treatment strategies [15] by innovative computer-assisted technologies, the comparison of different intraoperative treatment strategies [16], the evaluation of surgical mistakes [17] and the evaluation of the application of surgical tools and devices [18] or the surgical assist systems [19]. Furthermore, the approaches feature early engineering developments for the control of semi-automated surgical tasks [20], as well as process engineering [21].
In the associated literature, there is only a very limited number of approaches that explicitly deal with the acquisition and modeling of surgical processes. MacKenzie et al. [22] presented a hierarchically organized model of a laparoscopic fundoplication procedure according to Nissen and Jannin et al. who described a method of modeling supratentatorial tumor removals in the area of neurosurgery [23, 24]. However, neither of these two considered the identification of surgical treatment profiles.
We present our feasibility study along with the measurement method that was used to model the surgical treatments and the strategy of computing surgeon–individual treatment profiles. To our best knowledge, a comparable approach is not available in the appertaining literature. Therefore, we believe that this work represents an important contribution to evidence-based surgery.
Methods and materials
Patient sample and participating surgeons
In the course of the presented study, 105 cataract procedures were measured on a detailed-to-the-second work step level. All procedures were performed by experienced surgeons in their daily routine. The assignment of the patients to the respective surgeons was performed by an assistant medical director during the entry examination under exclusively clinical aspects with regard to the anticipated complications during the intervention and the health-related general condition of the patients. The decision concerning in- or outpatient intervention was made on the basis of the clinical guidelines of the German Ophthalmologic Society [25]. Two surgeons (nos. 1 and 2) performed outpatient procedures, while the third surgeon (no. 3) performed inpatient procedures. A further selection of patients, for instance with regard to age, gender or severity of affliction, did not take place, as this was irrelevant to demonstrate the applicability of the method. Table 1 shows the patient characteristics.
Table 1.
Patient characteristics for the study
| Surgeon no. 1 | Surgeon no. 2 | Surgeon no. 3 | |
|---|---|---|---|
| No. of cases | 36 | 18 | 51 |
| Mean patient age | 70.1 ± 9.6 | 63.5 ± 13.3 | 73.7 ± 7.8 |
| Sex (m/f) | 14/22 | 9/9 | 22/29 |
| Treated eye (right/left) | 18/18 | 6/12 | 29/22 |
Measurement of surgical processes
The computation of the SiTP was a multi-stage approach comprising definitions, measurements and database computations (cp. Fig. 1, left-hand side). An overview of relevant terms used to explain these methods is presented in Table 2. The computation of SiTPs is based on a detailed measurement of surgical activities ([26], Fig. 1, right-hand side): for every single surgical work step relevant data needs to be collected. This data encompasses information on what is being done, who performs the work step, whereby the work step is performed (meaning which instrument is being used), where at the patient’s body the work step is carried out and when it is performed. The collection of all surgical activities that were measured during the treatment of one patient is termed (patient-) individual Surgical Process Model (iSPM).
Fig. 1.
Stages for the measurement, computation, and analysis of surgeon–individual treatment profiles (SiTP, left-hand side) and principle of the data representation in the individual Surgical Process Model (iSPM, right-hand side)
Table 2.
Terms and definitions related to the method of measuring surgeon–individual treatment profiles
| Term | Definition |
|---|---|
| Surgical Process (SP) | Surgical procedure, performed at one specific patient |
| Surgical Process Model (SPM) | General term for a computer model of a surgical procedure course |
| Activity | Representation of a surgical work step in the Surgical Process Model |
| Individual Surgical Process Model (iSPM) | Computer model of a surgical procedure course |
| Generic Surgical Process Model (gSPM) | Statistical averaged computer model of multiple surgical procedure courses |
| Surgeon–individual treatment profile (SiTP) | gSPM that was computed for a number of patients that were treated by the same surgeon |
At the beginning of the measurement, the clinical nomenclature needs to be defined. It comprises the names for the surgical phases of the procedure type, the surgical instruments and appliances needed, and the anatomical and pathological structures affected. This was performed jointly by a trained medical observer and an experienced surgeon from the department.
The terms of the clinical nomenclature were used to configure the measurement software, the Surgical Workflow Editor [26, 27]. The Surgical Workflow Editor software was used by an observer to measure iSPMs (cp. Fig. 2). It is able to describe the Surgical Process in detail and to the split second. During the actual intervention, the observer ‘translates’ the observed work steps with the help of the software into a machine-readable format. The start of each beginning interventional phase is tagged and the data concerning who, what, whereby and where is being collected (cp. Table 3). The information concerning the when—the temporal information concerning the starting point and the breakpoint—is gathered automatically by the software.
Fig. 2.
Screenshot of the tool kit for data acquisition: the Surgical Workflow Editor
Table 3.
Examples of activity descriptions for surgical work steps
| Example activity 1 | Example activity 2 | |
|---|---|---|
| Who | Surgeon with right hand | Surgeon with right hand |
| What | Hydrodissection | Wash |
| Whereby | Sauter cannula | Sprinkler cannula |
| Where | Cortex | Conjunctiva |
| When | 00:05:30–00:06:10 | 00:02:30–00:02:40 |
All cataract procedures of the feasibility study were recorded by a trained medical student as observer. The observer had to complete a substantial training program before he started taking measurements. This program encompassed the dealing with the software and with the information concerning the procedure itself, such as the involved terms of the nomenclature. During data acquisition, the observer was present in the operating room during the actual intervention and operated the software on a touch screen tablet PC. The so recorded iSPMs were stored on the tablet PCs in machine-readable eXtensible Markup Language (XML) format.
Generation of surgeon–individual treatment profiles and statistical analysis
After the completion of the data acquisition, the iSPMs were transferred into a database for further processing. The activities in the iSPMs were grouped according their association to one of the surgical phases (cp. Table 4). Subsequently, a generic Surgical Process Model was computed for each of the surgeons and each of the surgical phases as surgeon–individual treatment profile. This gSPM contained the number and the average performance times of each activity and a probability for the following surgical activity. Finally, the gSPMs were filtered to delete infrequently occurring activities. The strategy has been described more specifically in Neumuth et al. [16].
Table 4.
Interventional core phases and their definitions
| Phase | Definition |
|---|---|
| Opening of the lens bag by rhexis cannula | From first paracentesis until end of material excision |
| Cataract removal | Form hydrodissection until end of irrigation/aspiration of lens cortex |
| Posterior chamber intraocular lens implantation (PC-IOL) | From incision widening until beginning of irrigation/aspiration of Healon® |
| Removal of Healon® | Irrigation/aspiration of Healon® from anterior chamber |
To assess the differences between the SiTPs of the three surgeons, we calculated means and standard deviations for the occurrence number and the mean duration of each activity as well as for the probabilities. A statistical analysis using Bonferroni tests with a significance level of α = 0.05 was performed with the help of the statistics software SPSS [28] to check the means for statistical significance.
Results
The cut–suture times showed averaged time spans of 00:23:01 ± 00:11:59 for surgeon no. 1, 00:30:03 ± 00:20:57 for surgeon no. 2 and 00:16:01 ± 00:04:39 for surgeon no. 3 (cp. Table 5). The surgical core phase following the conclusion of the preparation, until the end of the Healon® removal, was 00:15:17 ± 00:12:20 for surgeon no. 1, 00:22:17 ± 00:21:54 for surgeon no. 2 and 00:09:50 ± 00:03:22 for surgeon no. 3. The differences were not statistically significant, with the exception of the difference between surgeon nos. 2 and 3 (p < 0.001).
Table 5.
Cut–suture times and durations of the interventional phases (in hours/minutes/seconds and with mean ± standard deviation)
| mean ± SD | Surgeon no. 1 | Surgeon no. 2 | Surgeon no. 3 | Between subject effects | p values | ||
|---|---|---|---|---|---|---|---|
| nos. 1 and 2 | nos. 1 and 3 | nos. 2 and 3 | |||||
| Cut–suture time | 00:23:01 ± 00:11:59 | 00:30:03 ± 00:20:57 | 00:16:01 ± 00:04:39 | F = 12.4, p < 0.001 | p > 0.05 | p = 0.005 | p < 0.001 |
| Duration from begin of Opening the lens bag until end of removal of Healon® | 00:15:17 ± 00:12:20 | 00:22:17 ± 00:21:54 | 00:09:50 ± 00:03:22 | F = 8.4, p < 0.001 | p > 0.05 | p > 0.05 | p < 0.001 |
| Opening of the lens bag by rhexis cannula | 00:02:44 ± 00:01:08 | 00:02:57 ± 00:01:18 | 00:01:28 ± 00:00:28 | F = 26.6, p < 0.001 | p > 0.05 | p < 0.001 | p < 0.001 |
| Cataract removal | 00:09:34 ± 00:10:29 | 00:11:50 ± 00:08:41 | 00:05:42 ± 00:02:24 | F = 9.5, p < 0.001 | p = 0.05 | p > 0.05 | p < 0.001 |
| Posterior chamber intraocular lens implantation | 00:01:02 ± 00:01:17 | 00:00:50 ± 00:00:27 | 00:00:42 ± 00:00:29 | F = 1.5, p > 0.05 | p > 0.05 | p > 0.05 | p > 0.05 |
| Removal of Healon® | 00:01:10 ± 00:00:42 | 00:02:45 ± 00:03:51 | 00:01:37 ± 00:01:18 | F = 4.1, p = 0.02 | p = 0.02 | p > 0.05 | p > 0.05 |
For the surgical core phases, highly significant differences (p < 0.001) were determined; for the phase ‘Opening of the lens bag by rhexis cannula’, for instance, for the surgeon nos. 1 and 3, as well as for nos. 2 and 3. The differences in duration of the ‘Cataract removal’ phase were highly significant (p < 0.001) concerning surgeon nos. 2 and 3, but only on a low significance level concerning surgeon nos. 1 and 2 (p = 0.05). The implantation of the lens showed no significant differences between the surgeons, while the removal of Healon® again showed a difference between surgeon nos. 1 and 2, but again on a low significance level (p = 0.02).
Figure 3 shows the individual progression course of the Surgical Process for each of the three surgeons for the interventional phase ‘Opening of the lens bag by rhexis cannula’. For a higher lucidity, all activity sequences occurring with a probability of less than 10% for all three surgeons were filtered. For the same reason, activity sequences with a probability of occurrence of more than 40% for all three surgeons were highlighted as main path by bold lines. For activities thus highlighted, the durations of the work steps have been computed exemplarily in Table 5.
Fig. 3.
Visualization of the surgeon–individual treatment profiles (SiTPs) as generic Surgical Process Models of the three surgeons for the intervention phase ‘Opening of the lens bag by rhexis cannula’. The graph shows the most frequently measured activities. The probability of on activity following another one is indicated as percentage for each surgeon with the labels on the edges
Table 6 shows the average number of occurrences of a surgical activity during the phase and its average cumulated execution time. It turned out that the activity ‘Paracentesis with right hand’ was performed significantly more often by surgeon nos. 3 (p < 0.001) and 2 (p < 0.001) than by surgeon no. 1. However, the differences in averaged total execution times were not statistically significant.
Table 6.
Average performance frequencies of surgical activities and average total performance times for each surgeon
| ID | Activity | Surgeon no. 1 | Surgeon no. 2 | Surgeon no. 3 | p values | Surgeon no. 1 | Surgeon no. 2 | Surgeon no. 3 | p values | ||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| nos. 1 and 2 | nos. 1 and 3 | nos. 2 and 3 | nos. 1 and 2 | nos. 1 and 3 | nos. 2 and 3 | ||||||||
| 1 | Right hand | 0.53 ± 0.51 | 0.89 ± 0.32 | 1.00 ± 0.00 | p = 0.001 | p < 0.001 | p > 0.05 | 00:00:17 ± 00:00:30 | 00:00:10 ± 00:00:08 | 00:00:06 ± 00:00:01 | p > 0.05 | p = 0.01 | p > 0.05 |
| Paracentesis | |||||||||||||
| Paracentesis knife | |||||||||||||
| Cornea | |||||||||||||
| 2 | Right hand | 1.17 ± 0.45 | 1.50 ± 0.92 | 1.04 ± 0.34 | p > 0.05 | p > 0.05 | p = 0.005 | 00:00:06 ± 00:00:02 | 00:00:10 ± 00:00:05 | 00:00:04 ± 00:00:01 | p < 0.001 | p = 0.008 | p < 0.001 |
| Inject | |||||||||||||
| Healon® | |||||||||||||
| Chamber ant | |||||||||||||
| 3 | Right hand | 1.11 ± 0.52 | 1.28 ± 0.83 | 1.02 ± 0.24 | p > 0.05 | p > 0.05 | p > 0.05 | 00:01:15 ± 00:00:24 | 00:01:03 ± 00:00:23 | 00:00:34 ± 00:00:11 | p > 0.05 | p < 0.001 | p < 0.001 |
| Capsulorhexis | |||||||||||||
| Rhexis cannula | |||||||||||||
| Capsulalentis | |||||||||||||
| 4 | Right hand | 1.03 ± 0.17 | 1.00 ± 0.00 | 1.02 ± 0.24 | p > 0.05 | p > 0.05 | p > 0.05 | 00:00:05 ± 00:00:01 | 00:00:04 ± 00:00:01 | 00:00:03 ± 00:00:01 | p > 0.05 | p < 0.001 | p > 0.05 |
| Cut | |||||||||||||
| Lancet clear cut | |||||||||||||
| Cornea | |||||||||||||
| 5 | Right hand | 1.14 ± 0.35 | 1.06 ± 0.24 | 1.18 ± 0.43 | p > 0.05 | p > 0.05 | p > 0.05 | 00:00:06 ± 00:00:04 | 00:00:07 ± 00:00:03 | 00:00:05 ± 00:00:04 | p > 0.05 | p > 0.05 | p > 0.05 |
| Excision material | |||||||||||||
| Utrata’s tweezers | |||||||||||||
| Capsulalentis | |||||||||||||
| 6 | Right hand | 0.17 ± 0.38 | 0.44 ± 0.62 | 0.00 ± 0.00 | p = 0.01 | p > 0.05 | p < 0.001 | 00:00:06 ± 00:00:02 | 00:00:07 ± 00:00:03 | n.a. | p > 0.05 | - | - |
| Inject | |||||||||||||
| Vision blue® | |||||||||||||
| Chamber ant | |||||||||||||
| 7 | Right hand | 0.22 ± 0.59 | 0.33 ± 0.49 | 0.04 ± 0.28 | p > 0.05 | p > 0.05 | p > 0.05 | 00:00:20 ± 00:00:05 | 00:00:15 ± 00:00:07 | 00:00:16 ± .00:00:00 | p > 0.05 | p > 0.05 | p > 0.05 |
| Irrigate | |||||||||||||
| Sauter cannula | |||||||||||||
| Chamber ant | |||||||||||||
| 8 | Both hands | 0.47 ± 0.51 | 0.11 ± 0.32 | 0.00 ± 0.00 | p = 0.001 | p < 0.001 | p > 0.05 | 00:00:13 ± 00:00:20 | 00:00:08 ± 00:00:01 | n.a. | p > 0.05 | – | – |
| Paracentesis | |||||||||||||
| Paracentesis knife | |||||||||||||
| Cornea | |||||||||||||
| 9 | Left hand | 1.42 ± 0.69 | 1.83 ± 1.15 | 0.27 ± 0.49 | p > 0.05 | p < 0.001 | p < 0.001 | 00:01:42 ± 00:00:29 | 00:01:48 ± 00:00:52 | 00:00:37 ± 00:00:21 | p > 0.05 | p < 0.001 | p < 0.001 |
| Hold | |||||||||||||
| Colibri tweezers | |||||||||||||
| Bulbus oculi | |||||||||||||
| 10 | Left hand | 0.00 ± 0.00 | 0.00 ± 0.00 | 0.92 ± 0.59 | p > 0.05 | p < 0.001 | p < 0.001 | n.a. | n.a. | 00:00:48 ± 00:00:13 | – | – | – |
| Hold | |||||||||||||
| Micro spatula | |||||||||||||
| Bulbus oculi | |||||||||||||
The number of the occurrences of activity ‘Capsulorhexis with right hand’, on the other hand, was not significantly different; while the averaged cumulated durations were 00:01:15 ± 00:00:24 (surgeon no. 1), 00:01:03 ± 00:00:23 (surgeon no. 2) and 00:00:34 ± 00:00:11 (surgeon no. 3) and therefore significantly different between surgeon nos. 1 and 3 and nos. 2 and 3, respectively (p < 0.001). The utilization of the different surgical instruments, micro spatula and colibri tweezers to hold the Bulbus oculi, by surgeon no. 3 is also visible in the analysis’ results.
The probability of sequence for the core activities of the surgical workflow and the respective differences in the SiTPs are represented in Table 7 and Fig. 3. While most of the activity sequences in the SiTPs were not highly significantly different, the probability for the occurrence of the sequence ‘Paracentesis with right hand’ to ‘Injection of Healon®’ was highly significant for surgeon nos. 1 and 3 on the one hand (p < 0.001), and surgeon nos. 2 and 3 on the other (p < 0.004). The performance of both-handed paracenteses by surgeon nos. 1 and 2 is represented in the results as statistically significant differences in the SPMs compared to the surgeon no. 3.
Table 7.
Sequence probability for the work steps of the right hand (ideal progression course)
| ID | Start activity | Stop activity | Surgeon no. 1 | Surgeon no. 2 | Surgeon no. 3 | p values | ||
|---|---|---|---|---|---|---|---|---|
| nos. 1 and 2 | nos. 1 and 3 | nos. 2 and 3 | ||||||
| S-1 | START | Right hand | 0.53 ± 0.51 | 0.89 ± 0.32 | 1.00 ± 0.00 | p = 0.001 | p < 0.001 | p > 0.05 |
| Paracentesis | ||||||||
| Paracentesis knife | ||||||||
| Cornea | ||||||||
| S-8 | START | Both hands | 0.47 ± 0.51 | 0.11 ± 0.32 | 0.00 ± 0.00 | p = 0.001 | p < 0.001 | p > 0.05 |
| Paracentesis | ||||||||
| Paracentesis knife | ||||||||
| Cornea | ||||||||
| S-9 | START | Left hand | 0.47 ± 0.51 | 0.83 ± 0.38 | 0.24 ± 0.43 | p = 0.02 | p > 0.05 | p < 0.001 |
| Hold | ||||||||
| Colibri tweezers | ||||||||
| Bulbus oculi | ||||||||
| S-10 | START | Left hand | 0.00 ± 0.00 | 0.00 ± 0.00 | 0.67 ± 0.48 | p > 0.05 | p < 0.001 | p < 0.001 |
| Hold | ||||||||
| Micro spatula | ||||||||
| Bulbus oculi | ||||||||
| 1-2 | Right hand | Right hand | 0.42 ± 0.5 | 0.56 ± 0.51 | 0.92 ± 0.27 | p > 0.05 | p < 0.001 | p = 0.004 |
| Paracentesis | Inject | |||||||
| Paracentesis knife | Healon® | |||||||
| Cornea | Chamber ant | |||||||
| 1-6 | Right hand | Right hand | 0.03 ± 0.17 | 0.28 ± 0.46 | 0.00 ± 0.00 | p < 0.001 | p > 0.05 | p < 0.001 |
| Paracentesis | Inject | |||||||
| Paracentesis knife | Vision blue® | |||||||
| Cornea | Chamber ant | |||||||
| 2-3 | Right hand | Right hand | 0.84 ± 0.34 | 0.82 ± 0.38 | 0.81 ± 0.37 | p > 0.05 | p > 0.05 | p > 0.05 |
| Inject | Capsulorhexis | |||||||
| Healon® | Rhexis cannula | |||||||
| Chamber ant | Capsulalentis | |||||||
| 3-2 | Right hand | Right hand | 0.01 ± 0.04 | 0.10 ± 0.24 | 0.00 ± 0.00 | p = 0.004 | p > 0.05 | p = 0.001 |
| Capsulorhexis | Inject | |||||||
| Rhexis cannula | Healon® | |||||||
| Capsulalentis | Chamber ant | |||||||
| 3-4 | Right hand | Right hand | 0.94 ± 0.22 | 0.68 ± 0.4 | 0.90 ± 0.28 | p = 0.006 | p > 0.05 | p = 0.02 |
| Capsulorhexis | Cut | |||||||
| Rhexis cannula | Lancet clear cut | |||||||
| Capsulalentis | Cornea | |||||||
| 4-5 | Right hand | Right hand | 0.99 ± 0.08 | 1.00 ± 0.00 | 0.94 ± 0.22 | p > 0.05 | p > 0.05 | p > 0.05 |
| Cut | Excision material | |||||||
| Lancet clear cut | Utrata’s tweezers | |||||||
| Cornea | Capsulalentis | |||||||
| 5-5 | Right hand | Right hand | 0.07 ± 0.18 | 0.03 ± 0.12 | 0.10 ± 0.20 | p > 0.05 | p > 0.05 | p > 0.05 |
| Excision material | Excision material | |||||||
| Utrata’s tweezers | Utrata’s tweezers | |||||||
| Capsulalentis | Capsulalentis | |||||||
| 5-E | Right hand | END | 0.93 ± 0.18 | 0.97 ± 0.12 | 0.86 ± 0.27 | p > 0.05 | p > 0.05 | p > 0.05 |
| Excision material | ||||||||
| Utrata’s tweezers | ||||||||
| Capsulalentis | ||||||||
| 6-7 | Right hand | Right hand | 0.17 ± 0.38 | 0.25 ± 0.43 | 0.00 ± 0.00 | p > 0.05 | p = 0.02 | p = 0.005 |
| Inject | Irrigate | |||||||
| Vision blue® | Sauter cannula | |||||||
| Chamber ant | Chamber ant | |||||||
| 7-2 | Right hand | Right hand | 0.15 ± 0.35 | 0.33 ± 0.49 | 0.00 ± 0.00 | p > 0.05 | p > 0.05 | p < 0.001 |
| Irrigate | Inject | |||||||
| Sauter cannula | Healon® | |||||||
| Chamber ant | Chamber ant | |||||||
| 8-2 | Both hands | Right hand | 0.36 ± 0.49 | 0.06 ± 0.24 | 0.00 ± 0.00 | p = 0.002 | p < 0.001 | p > 0.05 |
| Paracentesis | Inject | |||||||
| Paracentesis knife | Healon® | |||||||
| Cornea | Chamber ant | |||||||
| 8-9 | Both hands | Left hand | 0.42 ± 0.50 | 0.11 ± 0.32 | 0.00 ± 0.00 | p = 0.004 | p < 0.001 | p > 0.05 |
| Paracentesis | Hold | |||||||
| Paracentesis knife | Colibri tweezers | |||||||
| Cornea | Bulbus oculi | |||||||
| 9-E | Left hand | END | 0.83 ± 0.27 | 0.71 ± 0.31 | 0.23 ± 0.42 | p > 0.05 | p < 0.001 | p < 0.001 |
| Hold | ||||||||
| Colibri tweezers | ||||||||
| Bulbus oculi | ||||||||
| 9-9 | Left hand | Left hand | 0.14 ± 0.25 | 0.25 ± 0.30 | 0.01 ± 0.07 | p > 0.05 | p = 0.009 | p < 0.001 |
| Hold | Hold | |||||||
| Colibri tweezers | Colibri tweezers | |||||||
| Bulbus oculi | Bulbus oculi | |||||||
| 10-E | Left hand | END | 0.00 ± 0.00 | 0.00 ± 0.00 | 0.70 ± 0.43 | p > 0.05 | p < 0.001 | p < 0.001 |
| Hold | ||||||||
| Micro spatula | ||||||||
| Bulbus oculi | ||||||||
Discussion
The identification of surgeon–individual treatment profiles supports the provision of an optimum treatment service of ophthalmological departments. While the performance of a complete homogeneous treatment service is not possible due to patient-specific characteristics, the ophthalmological department can promote to come close to this objective by an advanced training of their surgical staff members for frequently occurring treatments, such as cataract procedures. We demonstrated with our feasibility study that it possible to measure and compute SiTPs. Based on this, options for individual advanced training can be identified for each surgeon.
The results of the example phase demonstrated that it is possible to measure detailed quantitative information concerning the ascertainable criteria frequency, duration and sequence of surgical activities. With the help of the individual and generic Surgical Process Models as used in this work, treatment profiles could be derived for each surgeon.
As was shown by means of the example of the interventional phases, a selective appraisal of different time spans is possible. For instance, the differences in cut–suture times (p < 0.001) and in the overall durations of the single core phases (p = 0.001) from surgeon nos. 1 to 2 and 3 were highly significant. A further investigation has shown that these differences are mainly due to the different time spans of the phases ‘Opening of the lens bag by rhexis cannula’ (p < 0.001) and ‘Cataract removal’ (p < 0.002).
The treatment profile of surgeon no. 3 appeared to be distinctly different from those of the other two surgeons. With him, work step repetitions occurred less frequently, for instance concerning the work step ‘Inject Healon® with right hand in chamber anterior’. His work steps had a shorter duration, for instance ‘Capsulorhexis with rhexis cannula’. Furthermore, he had a distinct preference concerning the instrument used to ‘Hold bulbus oculi’ as compared to surgeon nos. 1 and 2.
It was furthermore possible to identify the differences in the activity sequences in the treatment profiles, as shown at the example of the sequence ‘Paracentesis with right hand’ and ‘Injection of Healon®’. Surgeon no. 3 had a probability of occurrence of 92% concerning these activity sequences on his primary route, while the profile of surgeon no. 1 showed 56% and the profile of surgeon no. 1 showed 42% probability. Here, the great differences between the three surgeons can be explained by the performance of additional surgical work steps, namely ‘Inject Vision Blue®’ and ‘Irrigate with sauter cannula’ as performed by surgeon nos. 1 and 2. We cannot make a detailed analysis whether this surgical decision was necessary for the individual patient.
From the presented results, a concrete recommendation for the advanced surgical training can be derived. Firstly, a profound exchange of experience would be very useful. Surgeon no. 3, for instance, had the least repetitions and usually shortest performance times. Also, a training of the work step ‘Paracentesis with right hand’ could lead to an amelioration of results.
The presented data acquisition method allows for a detailed description of surgical processes and has been validated profoundly in previous studies [26]. At the same time, the validity of the presented results was secured by the extensive training and instruction of the observers. In previous studies, it has already been evaluated that it is possible to compute a gSPM from a number of iSPMs. This strategy allowed for representative computing of a treatment profile by preserving the variability of the surgical processes on the one hand and by eliminating infrequently occurring surgical activities on the other hand. It was also shown that gSPM generated from iSPMs lead to the procedure course that was recommended by clinical guidelines [16].
An improvement of the presented study might be the control of the allocation of patients to the single surgeons. In this study, this allocation was not performed on the basis of age, gender, or severity of affliction, but it was regarded as randomized, because it can be assumed that the surgeon cannot chose his patients in his normal work either. Nevertheless, an influence of different surgical severity cannot be excluded. The examination of these influences remains to be done in future clinical studies. Furthermore, the setting of one surgeon performing inpatient procedures and two surgeons performing outpatient procedures was chosen on purpose to demonstrate the feasibility of the approach to compare treatment profiles between surgeons performing the same strategy (surgeon nos. 1 vs. 2) and surgeons performing different strategies (surgeon nos. 1 resp. 2 vs. 3). Finally, we have intentionally refrained from discussing the reasons for the determined temporal differences in detail, because after the demonstration of the feasibility of the generation of surgeon–individual treatment profiles, the design of clinical studies can be featured.
Conclusion
The success of an ophthalmological department mainly depends on the capabilities of the surgical staff. We showed that it is possible to identify SiTP to promote advanced training of surgical staff and therefore to support optimal patient treatment service. By means of the feasibility study, it was shown that detailed profiles could be gathered with the help of Surgical Process Modeling that provides an exact, validated and objective decision base for the support of surgical teaching in the realm of evidence-based eye surgery.
Further extensions of treatment profiles are conceivable based on the presented results of this study. The computation of treatment profiles for using different surgical strategies or different surgical instruments is possible due to the availability of the method. These assessments could be done in cooperation with ophthalmologic societies to identify the best-practice know-how for the optimal patient care.
Acknowledgements
The authors thank the team that supported the performance of the study and the preparation of the article at the Innovation Center Computer Assisted Surgery, University of Leipzig: Caroline Elzner and Michael Thiele. The authors also thank the surgeons that were subject to this study for their willingness to participate. ICCAS is funded by the German Federal Ministry of Education and Research (BMBF) and the Saxon Ministry of Science and Fine Arts (SMWK) in the scope of the Unternehmen Region with the grant numbers 03 ZIK 031 and 03 ZIK 032 and by funds of the European Regional Development Fund (ERDF) and the state of Saxony within the frame of measures to support the technology sector.
References
- 1.Ezra DG, Chandra A, Okhravi N, Sullivan P, McDonnell P, Lee J. Higher surgical training in ophthalmology: trends in cumulative surgical experience 1993–2008. Eye [Internet]. 2010 Apr 30 [zitiert 2010 Juli 5];Available from: http://dx.doi.org/10.1038/eye.2010.54. [DOI] [PubMed]
- 2.Ayanniyi AA, Adepoju FG, Owoeye JF. Trainee ophthalmologists’ opinions on ways to improve cataract surgical rate. Ann Afr Med. 2009;8(4):276–80. doi: 10.4103/1596-3519.59585. [DOI] [PubMed] [Google Scholar]
- 3.AHRQ. Agency for Health Care Research and Quality: National Guideline Clearinghouse [Internet]. 2008;Available from: http://www.guideline.gov.
- 4.AWMF-Arbeitsgemeinschaft der Wissenschaftlichen Medizinischen Fachgesellschaften e.V. Science-based Guidelines for Diagnostics and Therapy issued by the Scientific Medical Societies in Germany [Internet]. 2010 [zitiert 2010 Juni 29];Available from: http://www.awmf-leitlinien.de/.
- 5.Henderson B, Kim J, Golnik K, Oetting T, Lee A, Volpe N, et al. Evaluation of the virtual mentor cataract training program. Ophthalmology. 2010;117(2):253–8. doi: 10.1016/j.ophtha.2009.07.009. [DOI] [PubMed] [Google Scholar]
- 6.Prakash G, Jhanji V, Sharma N, Gupta K, Titiyal JS, Vajpayee RB. Assessment of perceived difficulties by residents in performing routine steps in phacoemulsification surgery and in managing complications. Can J Ophthalmol. 2009;44(3):284–7. doi: 10.3129/i09-051. [DOI] [PubMed] [Google Scholar]
- 7.Rogers GM, Oetting TA, Lee AG, Grignon C, Greenlee E, Johnson AT, et al. Impact of a structured surgical curriculum on ophthalmic resident cataract surgery complication rates. J Cataract Refract Surg. 2009;35(11):1956–60. doi: 10.1016/j.jcrs.2009.05.046. [DOI] [PubMed] [Google Scholar]
- 8.Hashimoto C, Kurosaka D, Uetsuki Y. Teaching continuous curvilinear capsulorhexis using a postmortem pig eye with simulated cataract. J Cataract Refract Surg. 2001;27(6):814–6. doi: 10.1016/S0886-3350(00)00728-8. [DOI] [PubMed] [Google Scholar]
- 9.Figueira E, Wang L, Brown T, Masselos K, Pandya V, Dauber S, et al. The grape: an appropriate model for continuous curvilinear capsulorhexis. J Cataract Refract Surg. 2008;34(9):1610–1. doi: 10.1016/j.jcrs.2008.04.049. [DOI] [PubMed] [Google Scholar]
- 10.Fisher JB, Binenbaum G, Tapino P, Volpe NJ. Development and face and content validity of an eye surgical skills assessment test for ophthalmology residents. Ophthalmology. 2006;113(12):2364–70. doi: 10.1016/j.ophtha.2006.08.018. [DOI] [PubMed] [Google Scholar]
- 11.Cremers SL, Ciolino JB, Ferrufino-Ponce ZK, Henderson BA. Objective Assessment of Skills in Intraocular Surgery (OASIS) Ophthalmology. 2005;112(7):1236–41. doi: 10.1016/j.ophtha.2005.01.045. [DOI] [PubMed] [Google Scholar]
- 12.Taylor JB, Binenbaum G, Tapino P, Volpe NJ. Microsurgical lab testing is a reliable method for assessing ophthalmology residents’ surgical skills. Br J Ophthalmol. 2007;91(12):1691–4. doi: 10.1136/bjo.2007.123083. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Saleh GM, Gauba V, Mitra A, Litwin AS, Chung AKK, Benjamin L. Objective structured assessment of cataract surgical skill. Arch Ophthalmol. 2007;125(3):363–6. doi: 10.1001/archopht.125.3.363. [DOI] [PubMed] [Google Scholar]
- 14.Bhogal MM, Angunawela RI, Little BC. Use of low-cost video recording device in reflective practice in cataract surgery. J Cataract Refract Surg. 2010;36(4):542–6. doi: 10.1016/j.jcrs.2009.10.053. [DOI] [PubMed] [Google Scholar]
- 15.Neumuth T, Trantakis C, Riffaud L, Strauss G, Meixensberger J, Burgert O. Assessment of technical needs for surgical equipment by Surgical Process Models. Minim Invasive Ther Allied Technol. 2009;18(6):841–9. doi: 10.3109/13645700903384484. [DOI] [PubMed] [Google Scholar]
- 16.Neumuth T, Jannin P, Schlomberg J, Meixensberger J, Wiedemann P, Burgert O. Analysis of surgical intervention populations using generic surgical process models. Int J CARS. 2010;6:59–71. doi: 10.1007/s11548-010-0475-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Malik R, White P, Macewen C. Using human reliability analysis to detect surgical error in endoscopic DCR surgery. Clin Otolaryngol Allied Sci. 2003;28(5):456–60. doi: 10.1046/j.1365-2273.2003.00745.x. [DOI] [PubMed] [Google Scholar]
- 18.Mehta N, Haluck R, Frecker M, Snyder A. Sequence and task analysis of instrument use in common laparoscopic procedures. Surg Endosc. 2002;16(2):280–5. doi: 10.1007/s004640080009. [DOI] [PubMed] [Google Scholar]
- 19.Strauss G, Fischer M, Meixensberger J, Falk V, Trantakis C, Winkler D, et al. Workflow analysis to assess the efficiency of intraoperative technology using the example of functional endoscopic sinus surgery. HNO. 2006;54(7):528–35. doi: 10.1007/s00106-005-1345-8. [DOI] [PubMed] [Google Scholar]
- 20.Münchenberg JE, Brief J, Raczkowsky J, Wörn H, Hassfeld S, Mühling J. Operation planning of robot supported surgical Interventions. In: Proceedings of IEEE/RSJ International Conference on Intelligent Robots and Systems. IEEE Computer Society; 2001. p. 547–52.
- 21.Casaletto JA, Rajaratnam V. Surgical process re-engineering: carpal tunnel decompression—a model. Hand Surg. 2004;9(1):19–27. doi: 10.1142/S0218810404002066. [DOI] [PubMed] [Google Scholar]
- 22.MacKenzie C, Ibbotson A, Cao C, Lomax A. Hierarchical decomposition of laparoscopic surgery: a human factors approach to investigating the operating room environment. Min Invas Ther All Technol. 2001;10(3):121–8. doi: 10.1080/136457001753192222. [DOI] [PubMed] [Google Scholar]
- 23.Jannin P, Raimbault M, Morandi X, Riffaud L, Gibaud B. Model of surgical procedures for multimodal image-guided neurosurgery. Comput Aided Surg. 2003;8(2):98–106. doi: 10.3109/10929080309146044. [DOI] [PubMed] [Google Scholar]
- 24.Jannin P, Morandi X. Surgical models for computer-assisted neurosurgery. Neuroimage. 2007;37(3):783–91. doi: 10.1016/j.neuroimage.2007.05.034. [DOI] [PubMed] [Google Scholar]
- 25.DOG. Veröffentlichungen der Deutschen Ophthalmologischen Gesellschaft e.V. [Internet]. Deutsche Ophthalomoligsche Gesellschaft e.V. 2009;Available from: http://www.dog.org/publikationen/index.html.
- 26.Neumuth T, Jannin P, Strauss G, Meixensberger J, Burgert O. Validation of knowledge acquisition for surgical process models. J Am Med Inform Assoc. 2009;16(1):72–80. doi: 10.1197/jamia.M2748. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Neumuth T, Durstewitz N, Fischer M, Strauß G, Dietz A, Meixensberger J, et al. Structured recording of intraoperative surgical workflows. In: Horii S, Ratib O, herausgeber. Bellingham, WA: 2006. p. CID 61450A.
- 28.SPSS Inc. SPSS 15.0 [Internet]. Chicago: 2009. Available from: http://www.spss.com