|Year : 2021 | Volume
| Issue : 1 | Page : 142-147
Optimal time management on linear accelerator: A logical model to reduce patient waiting time
Avinash H Udayashankar1, Shibina Noorjahan1, Nirmala Srikantia1, Ravindra Babu Kamunuri1, Niranjan Immanuel1, Sandeep Muzumder1, SH Chandramouli2, Mazhar H Shariff1
1 Department of Radiation Oncology, St. John's Medical College Hospital, St. John's National Academy of Health Sciences, Bengaluru, Karnataka, India
2 Department of Radiotherapy, National Centre for Cancer Care and Research, Doha, Qatar
|Date of Submission||10-Dec-2018|
|Date of Decision||14-Feb-2019|
|Date of Acceptance||13-May-2019|
|Date of Web Publication||01-Nov-2019|
Avinash H Udayashankar
Department of Radiation Oncology, St John's Medical College Hospital, St John's National Academy of Health Sciences, Sarjapur Road, Bengaluru - 560 034, Karnataka
Source of Support: None, Conflict of Interest: None
Introdction: Optimal time management is of utmost importance in the radiotherapy department. Inappropriate allocation of time slots leads to prolonged waiting times and decreased patient satisfaction during external beam radiotherapy. The present study tests a logical model to improve the waiting time for the patients.
Materials and Methods: The treatment time, waiting time, and causes of delay were studied from November 4, 2014, to July 24, 2015. New rules were framed for treatment slot allocation from December 26, 2014. The treatment slots were classified based on the treatment technology (three-dimensional conformal radiotherapy and intensity-modulated radiotherapy) with inclusion of “buffer slots” and patient education. The results were compared before and after rules.
Results: A total of 1032 time slots were analyzed, of which 225 “before rules” and 807 “after rules,” respectively. There was a significant reduction in the average waiting time for treatment in on-time patients (median [interquartile range (IQR)] of 25.2 min [31.75] vs. 3 min [3.5]; P< 0.00001) as well as in late-coming patients (median [IQR] of 38.2 min [13.795] vs. 21.11 min [12.75]; P= 0.00006). 59.7% (71 patients) of the treatment was delayed “before rules” as opposed to 32.2% (137 patients) “after rules” in on-time patients. Due to better patient education, there was a significant improvement in the patient punctuality toward the allotted time.
Conclusion: The treatment slots classified based on the teletherapy technique with buffer slots, and patient education helps in better time management on linear accelerator. This methodology significantly reduces waiting time and thereby the number of patients having delay in the treatment.
Keywords: Appointments and schedules, external beam radiotherapy, patient education, patient satisfaction, patient waiting time, time management
|How to cite this article:|
Udayashankar AH, Noorjahan S, Srikantia N, Kamunuri RB, Immanuel N, Muzumder S, Chandramouli S H, Shariff MH. Optimal time management on linear accelerator: A logical model to reduce patient waiting time. J Can Res Ther 2021;17:142-7
|How to cite this URL:|
Udayashankar AH, Noorjahan S, Srikantia N, Kamunuri RB, Immanuel N, Muzumder S, Chandramouli S H, Shariff MH. Optimal time management on linear accelerator: A logical model to reduce patient waiting time. J Can Res Ther [serial online] 2021 [cited 2021 Apr 17];17:142-7. Available from: https://www.cancerjournal.net/text.asp?2021/17/1/142/270100
| > Introduction|| |
External beam radiotherapy is the most important modality in the treatment of cancer patients. India is a low-middle income country which sees ≥1 million new cancer cases every year., As per the literature, at least five of ten patients diagnosed with cancer would require radiotherapy. Many radiotherapy centers in India treat 50–90 patients on each machine, working for 8–14 h in a day. In the process of treating huge number of patients every day, often patient satisfaction in terms of waiting time takes a back seat. Waiting time has one of the greatest influences on the overall satisfaction and can lead to psychological distress in patients.
As the number of patients treated on linear accelerator increases, it is challenging to maintain the time slots. Institutions often end up having huge backlog of cases daily, with large number of patients waiting in front of the linear accelerator room for their treatment. This adversely affects the patient satisfaction, doctor–patient relationship, and patient compliance toward treatment. Importance of patient satisfaction cannot be overemphasized in modern practice.
Increased waiting time indicates a poorly organized process and lack of concern toward the patients. Radiotherapy centers need to work out an appropriate strategy for optimal utilization of the time on the machine. In our center, we had divided the working hours of the machine into treatment slots of 10–20 min for each patient based on the technology used. We had consistently faced the problem of prolonged waiting times for the patients.
There are very few studies which have studied the waiting times for daily teletherapy treatment.,, Chan et al. have analyzed waiting times and give reasoning for the delay but do not test any method to reduce the waiting times. Wallis et al. and Joseph et al. have analyzed the waiting times and proposed a computer-based model to reduce the waiting time. Wallis et al. have tested the key performance indicator model and have succeeded in reducing the waiting times, but Joseph et al. have proposed a model based on machine learning but did not test the same. Here, we propose and test a logical model for the allocation of treatment slots for reducing the waiting time.
We hence conducted an audit of treatment time and waiting time for patients on linear accelerator to identify the reasons for increased waiting time. Moreover, later, we devised certain set of rules based on logical assumptions for the allocation of time slots to reduce the patient waiting time and continued the audit to verify the efficacy of the same. The present study reports the results of the audit.
| > Materials and Methods|| |
The study design was “clinical audit.” The audit was initiated by the Department of Radiation Oncology, St. John's Medical College Hospital, in November 2014 to identify the causes of delay in the treatment and prolonged waiting time for the patients receiving teletherapy. The initial audit was conducted two times in a week by a radiation oncologist and a radiotherapy technologist from November 4, 2014, to December 26, 2014.
During the initial audit, the treatment slots were allocated in arbitrarily without considering treatment technique [Figure 1]a.
|Figure 1: Treatment slot allocation and consequence of delay. The numbers in the inner yellow ring indicate the chronology of patient. (a) An example treatment slot allocation “Before rules.” (b) The consequence of delay in the treatment of patients “Before Rules.” The delayed arrival of the 2nd patient on the day resulted in chain of events which lead to delay of >20 min for the last patient [refer Appendix A for individual patient details]. (c) An example treatment slot allocation “After rules.” And (d) The consequence of delay in the treatment of patients “After rules.” It can be noted that the buffer slots are negating the effects of delay at the end of every hour|
Click here to view
Based on the observations made in the initial audit, on December 29, 2014, we devised the following set of rules for allotting the treatment slots for the teletherapy patients on linear accelerator:
- The treatment slots were classified based on the treatment technique (broadly two techniques, namely three-dimensional conformal radiotherapy [3DCRT] and intensity-modulated radiotherapy [IMRT]) [Figure 1]c
- The IMRT slots can be taken over for 3DCRT patients if excess of 3DCRT patients exist and vice versa
- A “buffer slot” of 10 min was introduced after every five patients treated with 3DCRT and three patients treated with IMRT [Figure 1]c
- Patients were educated regarding the time slots. The patients were informed to come maximum 30 min before the allotted time and always avoid delay of >10 min. This avoids the unnecessary crowd in the waiting room
- The allocation of the time slots is exclusively done by the radiotherapy technologists. The slots should be allotted based on the rules without yielding on to the patient requests
- Patients who come later than the allotted time should be made to wait till an available empty or buffer slot
- The patients who have pretreatment preparations such as bladder protocol must be given an earlier reporting time based on the time slot on the machine
- Admitted patients were also allocated time slots based on the treatment technique
- The new patients to be started on treatment should be given separate slots, not to collide with the patients on daily treatment
- The quality assurance, maintenance procedures, and cleaning must be scheduled before or after the treatment of all the patients.
The audit was continued twice a week on every Tuesday and Thursday, for the next 7 consecutive months from December 30, 2014, to July 23, 2015. For the data analysis and interpretation, we called the audit from November 4, 2014, to December 26, 2014, as “Before Rules” and the same from December 30, 2014, to July 23, 2015, as “After Rules.”
On each audit day, for every patient, the following parameters were recorded:
- Allotted time – defined as the time slot allotted for the patient
- Arrival time – defined as the time at which the patient reported to the radiotherapy technologist on that day
- Treatment time – defined as the time taken for the patient to enter the treatment room till he or she comes out. It includes entry of patient to the treatment room, treatment setup, verification of the setup by appropriate onboard imaging, delivery of treatment, and exit of the patient after the treatment
- Waiting time – defined as the time duration from the allotted time to start of the treatment time for patients who arrived on time. In patients who arrived earlier to their allotted time, the waiting time would be calculated from the allotted time, not from the arrival time. Waiting time for the patients who arrived later than the allotted time was defined as the time duration from the time they arrived till the start of the treatment time
- Technique of treatment (3DCRT or IMRT) and type of imaging done on that day (electronic portal imaging device [EPID] or cone beam computed tomography [CBCT]) were recorded.
Based on the above recorded data, total number of treated patients, number of patients treated with 3DCRT technique, number of patients treated with 3DCRT technique and on-board image verification (EPID/CBCT), number of patients treated with IMRT technique, number of patients treated with IMRT and on-board image verification (EPID/CBCT), number of patients who arrived on time and number of patients who arrived late, and number of admitted patients treated were calculated for each audit day.
We also calculated minimum, maximum, and average waiting time for patients who arrived on time and for the patients who arrived late and average treatment time for 3DCRT (with and without on-board image verification) and IMRT (with and without on-board image verification). All the recorded data before and after rules were compared.
The cause of delay analysis
We classified the patients who were on time and got delayed for the treatment based on the causes enumerated in [Table 1]. The percentage of patients who were delayed due to various reasons before and after the rules were calculated and compared.
|Table 1: The list of causes of delay in the treatment which were analyzed before and after new model|
Click here to view
Microsoft Excel (2016) was used for entering data and initial analysis. Statistical tests were done using R Foundation for Statistical Computing, Vienna, Austria. The treatment time for different techniques was expressed in mean (±standard deviation [SD]) before and after rules. The number of patients treated and waiting times were expressed in median (interquartile range [IQR]). A two-sample t-test was used to compare means, before and after rules, when the data were normally distributed. Nonparametric Mann–Whitney U-test was used for data that were not normally distributed. In either case, statistical significance was kept at P < 0.05. The causes of delay were expressed in percentages.
| > Results|| |
There were 15 audit days before rules and 57 days after rules. In total, during the entire audit period, 1286 treatments were done on the linear accelerator, 280 and 1006, before and after rules, respectively.
There was no statistically significant difference between the median number of patients treated per day before and after rules [Table 2]. There was a significant reduction in the average waiting time for treatment in on-time patients (median [IQR] of 25.2 (31.75) min vs. 3 [3.5] min; P < 0.00001). Both minimum and maximum waiting period was significantly reduced after rules. The waiting time for patients who came late also significantly reduced (median [IQR] of 38.2 (13.795) min vs. 21.11 (12.75) min; P = 0.00006) [Table 3].
|Table 2: The summary of number of patients who were treated before and after new rules|
Click here to view
|Table 3: Delay in the treatment for patients who arrive on time and late, before, and after rules|
Click here to view
There was no statistically significant difference between the average treatment time (mean ± SD, 15.06 min ± 1.104 before rules vs. 15.6 min ± 1.446 after rules; P = 0.129). However, there was a statistically significant difference between average treatment time for 3DCRT before and after rules (mean ± SD, 10.44 ± 0.921 vs. 12.87 ± 1.33; P < 0.0001) [Table 4].
A total of 1032 outpatient slots were analyzed for causes of delay, of which 225 “before rules” and 807 “after rules,” respectively. Out of those, only 52.9% (119 patients) and 52.5% (426 patients) came on time in before and after rules audit, respectively. However, 59.7% (71 patients) were delayed before rules as opposed to 32.2% (137 patients) after rules. Predominant cause of delay both before and after rules was “catching up with previous delay” with 57.7% and 54.7%, respectively [Table 5].
|Table 5: Comparison of causes for delay before and after rules, for the patients who came on or before their allotted time|
Click here to view
The percentage of patients coming late remained almost the same before and after rules (47.1% [106 patients] and 47.5% [381 patients]). However, the median delay in arrival of late patients reduced significantly after rules (median [IQR], 15 min (15) vs. 45 min (57.5); P < 0.0001).
| > Discussion|| |
Our study showed that the proposed system of time slot allocation significantly reduces the waiting time for the patients. The study is unique in terms of comparing a preexisting model with a new logical model of time allocation. We successfully optimized the treatment slots and improved the quality of care given for the patients.
In contrary to the usual belief, waiting times can happen in a smaller number of patients treated in a day as shown by our initial audit. Chan et al. and Joseph et al. have demonstrated waiting times in smaller number of patients than in our study. Waiting time is an integral component of the slot-based system in daily radiotherapy; there can only be measures taken to reduce the waiting times, but we cannot completely remove it out of the system., Hence, low-volume centers must also audit to find the patient waiting times and can use our model to reduce waiting times.
We divided the treatment slots based on the treatment technique because we observed that the treatment time for each of these two broad techniques was different. The average time taken to treat a patient with IMRT and 3DCRT was 20 min and 10 min, respectively; it was ideal to club the patients being treated with similar technologies together. Before and during the initial audit, we used to randomly allocate time slots for patients with 3DCRT as 10 min and IMRT as 20 min (2 times a 10 min slot). Hence, for adding a new IMRT patient, we need to complete two consecutive 3DCRT patients or one IMRT patient, which is rare. Hence, when an IMRT patient gets added to a 3DCRT slot in the old slot system, there used to be a confusion and rearrangement of the subsequent slots, consequential delay. This was the reason why we classified the slots based on the treatment technique. This kind of technique-based division helps in the allocation of time slots to a new patient.
If one patient's treatment is delayed, subsequent patient's treatment also gets delayed due to chain reaction since buffer slots were not available before rules [Figure 1]b and [Appendix A]. The time slots are subdivided on hourly basis, and hence, the waiting time gets negated at the end of each hour. This explains why even the patients who came late also had significantly less waiting time as compared to the initial audit. After the initial audit, we found that predominant cause of delay was due to “catching up with previous delay” (57.7%). This was similar to what was reported before by Chan et al. We hence introduced “buffer slots” of 10 min duration at periodic intervals to negate the treatment delays at the end of every hour [Figure 1]d. However, “Catching up with previous delay” remains the most common cause of delay in the audit after rules as well. There were two causes for this: (a) proportion of patients coming late were almost the same before and after rules (47.1% and 47.5%) and (b) the average treatment time for 3DCRT technique was more after rules than before rules. It also must be noted that we have reasoned any amount of delay (whether 1 min or 1 h). Even though the cause of delay remains the same, the quantum of the delay was significantly reduced after rules (median [IQR] of 25.2 (31.75) min vs. 3 (3.5) min; P < 0.00001 for before vs. after rules, respectively).
The average treatment time for 3DCRT technique increased by a factor of 2 min in the audit after rules, which is statistically significant. The probable reason why there was increase in the treatment time was due to variation in the patient characteristics treated and longer duration of assessment done (2 months vs. 7 months). If there was greater proportion of stretcher-bound and wheelchair-bound patients treated, the treatment time increases due to time taken for shifting of patients to and from the couch.
Another important component of change in the system is education of patients in improving punctuality. Williams et al. showed that patient unpunctuality is an important factor which increases the patient waiting time. Our study showed that duration of delayed arrival of late patients “after rules” significantly reduced. Hence, patients became more punctual after the intervention.
Although ours is a low-volume center which treats on an average of 20 cases per day, similar system if adapted in a higher volume centers can improve the time management on linear accelerator, thereby able to treat more patients with better patient satisfaction.
The strength of the present study is that it is a systematically collected data for a period of 9 months that compares two systems of treatment slot allocation and tests a logical system which was derived from the observations of the initial audit. Second, the study tests show an effective simple system of classification of time slots as opposed to sophisticated software for the treatment scheduling.
The drawbacks of our study are as follows: first, the audit was done only 2 days in a week, the events of machine breakdown and delayed start may have happened on other 3 days of the week; and hence, the study might have underestimated the waiting times. Second, this system needs to be tested in larger volume centers like those who treat ≥30 patients. Third, the individual patient characteristics could not be studied since there can be patients treated with 3DCRT technique taking more time than IMRT. Finally, there could be a factor of bias since the audit was done by the treating radiotherapy technologists only twice a week.
| > Conclusion|| |
Based on our study, we conclude that allotment of treatment slots based on the technique of teletherapy treatment with inclusion of buffer slots and adequate education of patients helps in better time management on linear accelerator. This technique significantly reduces waiting time and reduces the number of patients having delay in the treatment.
We would like to thank Dr. Tinku Thomas, Professor, Biostatistics, St. John's National Academy of Health sciences, Mrs. Laisamma Thomas, Department Secretary, Department of Radiation Oncology.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| > References|| |
Murthy NS, Chaudhry K, Rath GK. Burden of cancer and projections for 2016, Indian scenario: Gaps in the availability of radiotherapy treatment facilities. Asian Pac J Cancer Prev 2008;9:671-7.
Baskar R, Lee KA, Yeo R, Yeoh KW. Cancer and radiation therapy: Current advances and future directions. Int J Med Sci 2012;9:193-9.
Kannan V, Bajpai R. Conforming modern radiation oncology facilities to the irregular contours of the vast and varied nation of India. Int J Radiat Oncol Biol Phys 2016;94:645-51.
Famiglietti RM, Neal EC, Edwards TJ, Allen PK, Buchholz TA. Determinants of patient satisfaction during receipt of radiation therapy. Int J Radiat Oncol Biol Phys 2013;87:148-52.
Prakash B. Patient satisfaction. J Cutan Aesthet Surg 2010;3:151-5.
] [Full text]
Wallis A, Moretti D. PO-1023: Reducing waiting room times – A 5-year review of an in-house KPI tool. Radiother Oncol 2017;123:S564.
Chan K, Li W, Medlam G, Higgins J, Bolderston A, Yi Q, et al.
Investigating patient wait times for daily outpatient radiotherapy appointments (A single-centre study). J Med Imaging Radiat Sci 2010;41:145-51.
Joseph A, Hijal T, Kildea J, Hendren L, Herrera D. Predicting Waiting Times in Radiation Oncology Using Machine Learning. 16th
IEEE, International Conference on Machine Learning and Applications; 2017. p. 1024-9.
Williams KA, Chambers CG, Dada M, McLeod JC, Ulatowski JA. Patient punctuality and clinic performance: Observations from an academic-based private practice pain centre: A prospective quality improvement study. BMJ Open 2014;4:e004679.
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]