Identifying delays in the outpatient interventional radiology department: a quality improvement study

Abstract

Introduction

In the post‐COVID 19 environment, it has become increasingly important for healthcare services to optimise service delivery for the benefit of both patients and staff. The project purpose was to quantify and determine causes of throughput delays in a newly established outpatient angiography service in a public hospital setting.

Methods

This single‐centre study obtained quantitative and qualitative data for 81 consecutive outpatient interventional radiology (IR) examinations over a 3‐month period via survey and retrospective analysis of electronic medical records. Staff participating in data collection were able to record multiple causes for delay in a single case and were also able to include comments, allowing for more detailed descriptions of the delays that occurred.

Results

A total of 93 delay factors were identified in 73 of the 81 outpatient interventional examinations and grouped into six categories via thematic analysis. Availability of the IR room (40%), availability of the radiologist (28%) and insufficient documentation (18%) were identified as the most frequent causes for delay. Linear regression analysis showed that documentation (P = 0.0002) and room unavailability (P = 0.022) were independently associated with procedural starting delay.

Conclusion

Delays to the IR procedural start time occurred in 90% of cases (73/81). This study identified the causes for delays in outpatient interventional procedures. This information can be used to improve service delivery in IR departments.

This study identified the causes for delays in outpatient interventional procedures. Delays to the interventional radiology (IR) procedural start time occurred in 90% of cases (73/81). Availability of the IR room (40%), availability of the radiologist (28%) and insufficient documentation (18%) were identified as the most frequent causes for delay.

Introduction

Delays in the start time of radiology procedures and an ever‐changing schedule are typical occurrences in interventional radiology (IR) departments, leading to long working hours for staff, decreased patient satisfaction and exceptionally high rates of burnout. ¹ , ²

Wait times have been shown to have a strong association with the patient's perception of IR practices, with long wait‐times being a significant cause for complaints. ³ , ⁴ These challenges have been exemplified throughout the growth of our department, placing unprecedented demand upon IR resources and highlighting the need for improved workflow designs. One example of an improved workflow design involved the realignment of staff roles and responsibilities to create a new model of care. This was focused upon creating a structured process to clarify staff duties and encourage interprofessional communication. A single registered nurse (RN) managing the day‐to‐day operations in the department was also key to improving throughput. This saw a significant reduction in the time patients spent from the holding room to the procedure room. ⁵ In the post‐COVID environment, it has become vital for work practices to be optimised for the benefit of both patients and staff. ⁶

The installation of a new hybrid fluoroscopy and angiography suite at our hospital saw the expansion of the range of services provided to outpatients, inpatients and emergency patients. Services provided at our centre included a combination of barium‐based fluoroscopy tests, gynaecological investigations and IR procedures. Delays to the start time of IR procedures have been reported by staff involved in service delivery. A previous study reported that the causes of IR throughput delay were complex and multifactorial. These included factors such as the need for the physician to gain the patient's consent, lack of room availability and delays associated with laboratory testing and results. ⁷ It is hypothesised that a combination of delays was occurring in the workflow at our centre, including the addition of urgent cases, staffing availability and procedures taking longer than predicted. As these factors can compound, the demands upon staff can become significant.

Previous studies measuring throughput with the aim of reducing downtime in IR suites have highlighted the multidisciplinary approach required to achieve this. ⁵ , ⁸ Although the specific strategies employed are often varied, the approach requires involvement from all members of the department, including administration, nursing, radiographer and radiologist staff. Identification of the causes of throughput delay is therefore a vital first step that must be taken in order to improve service delivery.

Departmental improvements in previous studies have included the creation of a charge nurse role, improving the patient tracking board, defining team member's roles and responsibilities and revising staff rosters. ⁵ Another technique previously reported to save time is to map patient‐staff interactions and adjust workflows to ensure that steps that would normally occur in sequence can occur in parallel. ⁹ As such, the solution to improving hospital practices and the work‐life balance of staff members does not necessarily depend on increasing resources, but through efficient use of the resources that are available. ¹⁰

Previous studies have reported high rates of delays in the IR department, with 44–77% of cases commencing late prior to quality improvement (QI), and 28–44% of cases commencing late in the period following QI. ¹ , ⁷ , ¹¹ One study highlighted significant improvements achieved through the commencement of the first case of each day occurring on time. By adjusting the shift times of IR nurses, the proportion of cases starting late decreased from 100% to 44.9%. ¹²

The primary aim of this study was to measure the proportion of outpatient IR procedures that started on time. The study further aimed to identify causes for any delays. This would enable adjustments to service delivery that address the root causes of inefficiency.

Method

Ethical approval

The research project was assessed by the Monash Health research office and was exempted from ethics approval. Furthermore, the project was deemed consistent with the National Health and Medical Research Council Ethical Considerations in Quality Assurance and Evaluation Activities Guidelines. ¹³ Confidentiality was emphasised to all staff within the department in which this research project was performed. The staff responsible for data collection were provided with instructions and a template to ensure both patient and staff data would be collected in a manner that maintained confidentiality. All data were reported in aggregate, and no identifying information was included in any reports or presentations. Completed data sheets were stored within locked boxes in the Radiology Department. Data sheets will be stored for 7 years and subsequently permanently destroyed.

Data collection

This was a single centre study and incorporated both quantitative and qualitative data collection. This ensured that the research goals of comparing the scheduled to the start times of outpatient interventional procedures could be quantified and that causes of throughput delay could be identified.

The Medical Imaging Technologist (MIT) when rostered to the angiography suite was provided with written instructions and a survey was completed for consecutive outpatient IR procedures for a period of 3 months between June and September 2022.

Outpatients were selected for inclusion in the study as these patients are scheduled in advance according to a structured template. As a result, preparation and patient related factors could be accurately assessed as possible contributing factors to procedural delay. Inpatient and emergency IR studies were excluded from the study due to the ad hoc nature of scheduling and inability to quantify delays for urgent studies that have been accommodated at short notice. Surveys that were either incomplete or illegible were excluded from the study, as the reason for throughput delay could not be accurately determined.

The survey collected procedural information including the study date and time, type of procedure and the scheduled start/finish times. This information was used for the purposes of crosschecking examination data retrospectively. Identification of throughput delay and the factors contributing to the delay was performed via a tick list with the ability to provide additional qualitative data in the form of free text comments for further clarification. The selectable options on the tick list were based off a causal fishbone diagram for delayed throughput in an interventional service in research. ⁵ This enabled the research team to categorise the cause of throughput delay into the following factors; room availability, radiologist availability, nursing availability, documentation, patient‐related circumstances and other causes. As the delays were not ranked according to significance, and participating staff were able to choose more than one factor for a single case, the volume of factors causing delay exceeded the total number of cases.

Quantitative data associated with the patient arrival, schedule and procedural start and finish times were collected retrospectively to determine procedural length and this data was extracted by the researchers from the Philips Healthcare Radiology Informatics Solutions and Electronic Medical Records systems. These programs were used due to the specificity and accuracy of data. Continuous variables were summarised using medians and inter‐quartile ranges (IQR) and compared between procedure types using Kruskal–Wallis or Wilcoxon rank‐sum tests where appropriate. Categorical variables were reported using frequency counts and percentages. Univariate and multivariate analyses of length of delay were performed using linear regression with results reported as regression coefficients (standard errors (SE)) and an R ² statistic to indicate the amount of variation explained. Variables with a P < 0.20 on univariate regression were considered for inclusion in the multivariate model. All calculated P‐values were 2‐tailed and a P < 0.05 indicated statistical significance. Analyses were performed with SAS software version 9.4 (SAS institute, Cary, NC, USA).

Results

Data were collected for 81 cases in total. Table 1 presents the procedure type groupings. Table 2 presents the duration range of each procedure type and length of starting delay. Central line, urological and other studies had statistically significantly shorter procedural duration compared with embolisation and adrenal vein sampling studies (P = 0.001). The procedure type did not have any statistically significant impact on procedural starting delay (P = 0.95).

Table 1.

Procedure type groupings and the volume of cases.

Procedure type grouping	Examples	Number of delays	Percentage
Central line	PICC line insertion, port insertion, removal or exchange	54	67
Urological	Ureteric stent insertion, nephrostomy tube insertion, removal or exchange	13	16
Other	Rocket drain insertion, radiologically inserted gastrostomy, IVC filter insertion or removal	7	8
Adrenal vein sampling	Adrenal vein sampling	4	5
Embolisation	Ovarian vein embolisation, varicocoele embolisation	3	4

Table 2.

Procedure durations and length of starting delay.

Procedure type grouping	Procedure duration (minutes)		P‐value	Length of starting delay (minutes)		P‐value
	Median	IQR		Median	IQR
Central line	41.5	29–60	0.001	51	32–64	0.74
Urological	40	30–45		57	30–90
Other	37	29–55		50	30–60
Adrenal vein sampling	137	117–148		48	43–55
Embolisation	100	82–155		69	60–78

There were delays in 90% of cases (73/81). Of the 73 cases that encountered delays, 93 delay factors were identified. The volume of factors exceeded the total number of cases as staff were able to choose more than one delay factor for a single case. Table 3 presents the summary statistics for causes of IR procedural throughput delay.

Table 3.

Summary of delay factors for IR studies.

Delay factor	Examples	Number of delays	Percentage
Room unavailable	Previous case ran over time, further cleaning of room required	37	40
Radiologist unavailable	Radiologist in meeting, performing procedure in different modality, liaising with treating team, sick leave	26	28
Documentation	Uncertain patient preparation, lab results unavailable, previous imaging not available, consent form issues, ambiguous or poor‐quality request form requiring further clarification	15	16
Other	Training case with fellow or registrar, change in schedule due to urgent IP study	7	8
Nursing staff unavailable	Nursing team short of staff, involved in procedure in different modality	6	6
Patient circumstances	Patient running late or requiring use of bathroom	2	2

Of the delay factor categories considered, documentation and room unavailability were independently associated with increased procedural delay (Table 4).

Table 4.

Univariate and multivariate linear regression analysis of factors associated with interventional radiology procedural delay (minutes).

Cause of delay	Univariate			Multivariate
	Regression coefficient (SE)	P‐value	R 2 statistic (%)	Regression coefficient (SE)	P‐value
Documentation	30.88 (9.21)	0.001	13.7	36.15 (9.21)	0.0002
Room unavailable	10.32 (7.92)	0.197	2.3	17.45 (7.45)	0.022
Radiologist unavailable	−8.42 (8.31)	0.31	1.4
Nursing staff unavailable	−13.11 (14.50)	0.37	1.1
Other	10.63 (13.55)	0.44	0.9
Patient circumstances	−10.83 (24.51)	0.66	0.3

Discussion

There have been a number of studies aimed at identifying and addressing the delays that occur in the interventional suites of large academic hospitals with multiple modalities and suites available. ¹ , ⁷ , ⁸ , ¹⁰ , ¹¹ , ¹² , ¹⁴ This study demonstrated that similar challenges were faced at our 370 bed metropolitan hospital – which performed approximately 700 interventional angiography cases per year. Although this study only measured the delays occurring for outpatient examinations and did not include inpatient and emergency department examinations, there are nonetheless a number of comparisons that can be made.

Room availability

The most common cause of delay was the availability of the IR room, accounting for 40% (37/93). This is markedly higher than previously reported rates of 13% (93/718). ¹¹ A delay to the first case of the day typically resulted in cascading delays to subsequent cases. At our centre, the first case of the day is an inpatient video fluoroscopy speech study. Delays to ward transfers meant that these studies started later and extended beyond the allocated schedule into the timeslots reserved for the outpatient IR procedures. Inadequate preparation of the room, such as cleaning or inappropriate system setup were other factors that impacted on room availability.

IR cases at our centre are typically given 60‐minute appointments. On average, central line, urological and other IR procedure durations were between 37 and 41.5 min. This would suggest that the appointment length for these procedures is sufficient and does not impact on throughput delay. Conversely, embolisation studies and AVS studies had an average procedure length of 100 and 137 min respectively. This suggests that these types of procedures warrant an increased appointment duration of 150 min to allow for the wide variance in procedure length. The P‐value for procedure duration was calculated using the Kruskal–Wallis test, which showed that procedure duration differed by type (P = 0.001). Further analysis revealed that adrenal vein sampling and embolisation procures required a significantly longer length of time to be allocated when compared to all other procedure types (P < 0.05).

Documentation

A significant delay reported in previous literature is the need for physician involvement to clarify the order, obtain consent from the patient or provide pre‐sedation prior to the commencement of the case. Whilst this accounted for 16% (15/93) of delays in our study, in another study it accounted for 23.3% of delays by count, and 26.2% of delays by aggregate duration. ⁷

At our centre, every IR study underwent justification and approval by an IR consultant prior to scheduling. This process incorporated a review of the results of relevant pathology tests and previous imaging. Clear clinical guidelines relating to bleeding risk and anticoagulation were used to assist in peri‐procedural screening and management. The guidelines support appropriate coagulation and serology testing. This data was documented on the patient's record. Similarly, ambiguous or absent clinical details were discussed with the treating team prior to scheduling to ensure that the patient was booked in for the appropriate test. These processes are believed to contribute to a lower frequency of documentation‐related delays.

Patient‐related delays

Late arrival of the patient to the IR department has been reported as a frequent occurrence in previous literature, with 58.5% (297/508) of outpatients arriving late in one study. ¹¹ However, our study shows that this did not occur frequently, with only 2% (2/93) of delays occurring due to this reason during the sample period. At our facility, outpatients were instructed to arrive at least 1 h before their scheduled procedure time. An SMS confirmation was sent to all patients on the day prior to their appointment. Patients that did not confirm their booking received a follow‐up phone call from the administration team, which provided the patient with an opportunity to confirm, reschedule or cancel their appointment. This survey suggests that the booking process at our site was functioning well.

Staff availability

Availability of the radiologist was identified as a common cause for delays in our study, accounting for 28% (26/93) of delays during the sample period. This is somewhat consistent with previous research, showing that this accounted for 19.9% (101/508) of delays in outpatient interventional procedures. ¹¹ Changes to radiologist rostering, sick leave, conflicting meetings, phone calls and administrative duties were factors that contributed to radiologist availability. This supports the recommendation that non‐procedural duties should be delegated to alternative radiologists as a strategy for improving the workflow of outpatient procedures. ⁷ Nursing staff availability accounted for 6% of delays (6/93) and was typically associated with participation in interventions within other modalities or unplanned staff absences.

Other factors

Other factors accounted for 8% (7/93) of delays. The need to fit in an unscheduled study was considered an acceptable reason for delaying an outpatient IR as it was related to patient safety. Similarly, delays associated with registrar or fellow training were accepted given the rarity of certain examinations and the importance of maintaining a teaching culture.

As a result of this research, we recommend earlier starting times for MITs and nursing staff to allow adequate timing for planning, room preparation and communication to other relevant teams, such as ward based nurses or porters. Furthermore, cleaning staff could be provided with portable phones to facilitate timely clinical cleaning of the interventional suite. Future quality improvement plans relating to radiologist availability are currently being assessed, with the goal of reducing the administrative burden.

Limitations of the study

This was a single‐centre study and focused on a relatively small cohort of patients within the overall angiography and fluoroscopy service. Six examinations with missing or illegible data sheets were excluded from the study. These factors contributed to a relatively small sample size. Although a larger sample size may provide more reliable information, further complications may also be introduced due to the growth of service. Each delay factor was attributed equal weighting, although it is possible that certain causes of delays were more significant in individual cases. There was a possibility of the introduction of bias when attributing the factors causing throughput delay. Inpatient and emergency department patients were not included in the study as these examinations are typically scheduled in an ad hoc manner with flexibility as to the time the patient is brought to the department. Measuring the impact of updates to service delivery was not studied and is an area for further research.

Conclusion

Delays in the start time of outpatient IR procedures occurred in 90% of cases at our hospital. Availability of the IR room, availability of the radiologist and insufficient clinical documentation were identified as the most frequent causes of throughput delay. Insufficient clinical documentation and room unavailability were independently associated with procedural delay. The parameters of each IR service are unique, based on the population served and type of studies that are performed. A survey of this nature is recommended as a useful first step in identifying causes of IR throughput delay and optimising service delivery.

Funding Information

This research received no specific grant from any funding agency in the public, commercial or not‐for profit sectors.

Conflict of Interest

All authors declare that they have no conflicts of interest.

Ethics Statement

The research project was assessed by the Monash Health research office and was exempted from ethics approval. The project was deemed consistent with the National Health and Medical Research Council Ethical Considerations in Quality Assurance and Evaluation Activities Guidelines.

Data Availability Statement

The data that supports the findings of the study are available on request to the corresponding author.