Victorian Cancer Performance Monitoring Framework (VCPMF)

Transcription

1 Victorian Cancer Performance Monitoring Framework (VCPMF) Phase II Pilot Cancer Performance Indicator Results 2013 Data December 2016 Cancer Performance Indicator Results 2013 Data 1

2 Contents Invitation to review results 3 Introduction 4 - Indicator identification, selection & development 4 - Data collection statement 5 - Use of information 5 Data 6 - The Victorian Cancer Registry (VCR) 6 - Administrative data 6 - Data linkage 6 Cancer performance indicator results 8 - Performance Indicator 1: Emergency presentation and new cancer diagnosis 8 - Performance Indicator 6: Stage at diagnosis 13 - Performance Indicator 8: Reporting ECOG status performance 17 - Performance Indicator 13: Timeliness of initial treatment after cancer diagnosis 19 - Performance Indicator 16: Timeliness of adjuvant chemo for stage III colon cancer 34 - Performance Indicator 17: No. of lymph nodes examined during colon cancer surgery 36 Glossary 38 References 39 Appendices 40 - Appendix 1: Indicator selection and development flow chart - Appendix 2: ICS results by health service Tables: - Table 1: VCPMF Cancer Performance Indicators A Summary 4 - Table 2: Tumour stream definitions by International Classification of Disease (ICD) 7 - Table 3: ICS abbreviations 7 Cancer Performance Indicator Results 2013 Data 2

3 Invitation to review results 8 December 2016 Dear VICS Directors, Managers and Victorian cancer stakeholders Re: VCPMF Phase II Pilot: Cancer Performance Indicator Results 2013 Data We are pleased to provide the results for our initial suite of pilot cancer performance indicators for the 2013 data. The VCPMF Steering Group invites you to reflect on these results as part of a systematic process of analysis we are trialing for the Phase II pilot. The data are intended for use by each of the Integrated Cancer Services (ICS) and their clinical networks for internal review. The aim of the review is to assist with identifying aspects of cancer care in need of further investigation and quality improvement. The analysis, expertise and guidance provided by the ICS will play a central role in helping to close the loop with consumers, clinicians, executives, managers and other cancer stakeholders. We invite you to examine the data for what it might tell us about both the measures and what is happening in and across the ICS and their respective health services. In particular, we are interested in identifying any relevant trends and commonalities, variations and gaps. For example, is there a particular tumour stream or health service which shows a lower than expected proportion of patients receiving first treatment within 28 days? Does this suggest a need for further analysis at a local level? A couple of the pilot indicators such as stage distribution may become more useful as further data is added to demonstrate trends over time. It will be interesting to see trends in the state-wide average of 40% of stage III CRC patients receiving adjuvant intravenous chemotherapy within 8 weeks. A Guide to use of information (Guide) is the companion document to these results. The Guide provides an overview of the project s framework for quality and performance improvement, and charts the flow of information and deliverables across organisations involved in the analysis of the cancer data. For the benefit of the ICS, there is a simple step-by-step summary of the key tasks, timelines and deliverables (Appendix 3). We look forward to your engagement in this pilot program and any suggestions you might have about the presentation of the results and how this could be improved in the future. We also invite feedback on your experience of the state-wide reporting, clinical engagement, analysis and evaluation process. A link to an online survey to assist in this evaluation will be circulated shortly. The data analysis and written responses to results submitted by each of the ICS will be collated and processed to produce a state-wide VCPMF report on the findings. This will be made available in late March We plan to disseminate further results for the 2014 and 2015 data in February It is anticipated that these results will include data for some new cancer performance indicators as we seek to develop and expand the cancer performance monitoring and reporting. We hope you find the results interesting and that they are of both organisational and clinical value. Yours sincerely Dr Peter Briggs Chair, VCPMF Steering Group Clinical Director Southern Metropolitan Integrated Cancer Service Cancer Performance Indicator Results 2013 Data 3

4 Introduction The Victorian Cancer Performance Monitoring project is a joint undertaking between the Victorian Integrated Cancers Services (VICS) and Cancer Strategy and Development, Department of Health and Human Services (DHHS). The purpose of the project is to oversee the iterative development of a cancer performance monitoring framework for evaluating the quality and outcomes of cancer care. Phase I of the project developed a basic conceptual schema and operational model, undertook an indicator development process and embarked on initial capability testing of state-wide measures. This involved an extensive indicator development process and the initial capability testing of a suite of statewide measures. Table 1: VCPMF Cancer Performance Indicators A Summary No. Name OCP Step(s) 1. Emergency presentation and new cancer diagnosis Domain [2 & 3] Process, Outcome 6. Stage at diagnosis [3] Process, Outcome Tumour stream / cancer types Breast, Central Nervous System, Colorectal, Endocrine-Thyroid, Genito- Urinary, Gynaecological, Haematological, Head & Neck, Lung, Skin, Upper GI Breast, Colorectal 8. Reporting ECOG performance status [3] Process All tumour streams / cancer types 13. Timeliness of initial treatment after cancer diagnosis 16. Timeliness of adjuvant chemotherapy (intravenous) for stage III colon cancer 17. Number of lymph nodes examined during colon cancer surgery [3 & 4] Process Breast, Central Nervous System, Colorectal, Endocrine-Thyroid, Genito- Urinary, Gynaecological, Haematological, Head & Neck, Lung, Oesophago- Gastric, Skin [4] Process Colon cancer [4] Process, Outcome Colon cancer This document presents the pilot results of the initial tranche of six cancer performance indicators using 2013 data. Indicator identification, selection & development The indicator selection and development was a multi-staged process which drew on both evidencebased research and a range of alternative advice gathered through clinician engagement, stakeholder input and expert opinion. (See: Appendix 1 - Indicator selection and development flow chart) A key component of Phase I involved the identification of potential issues and areas for indicator development, and the specification and initial testing of measures. A set of seven criteria was established to guide the comparison, ranking and assessment of candidate indicator areas for selection by the VCPMF Steering Group. Cancer Performance Indicator Results 2013 Data 4

5 From an initial pool of around 500 potential indicator areas, six cancer performance indicators were shortlisted for development, specification and testing. An approved shortlist of 28 indicator areas was prioritised into three developmental waves for specification and testing. The indicator areas in each of the waves were reviewed for data collection feasibility and practicality. The final six shortlisted indicators for fast-track testing underwent a process of development and specification (Table 1.). Data collection statement The enclosed results relate to Victorians diagnosed in the calendar year These are identified via the Victorian Cancer Registry (VCR). Diagnoses in 2013 were chosen as this is the most recent year for which ICS of residence is known. The treatment data used in the results originate mainly from the Victorian Admitted Episodes Dataset (VAED), which is available up until July Cancer treatment outside of Victoria is not included. This will particularly affect the treatment data for patients in the Hume Regional Integrated Cancer Service who regularly receive their treatment in Albury, New South Wales. The Victorian Radiotherapy Minimum Data Set (VRMDS) is not used as it is not complete for the relevant period. We estimate a new data linkage will be available in early 2017 which will include Victorians diagnosed up to (and including) We expect the new data linkage will include the full VRMDS. Performance Monitoring Working Party A VCPMF Performance Monitoring Working Party has met on a weekly basis to attend to issues surrounding data collection and the presentation and reporting of results. The composition of the group included the Chair of the VCPMF Steering Group, a regional ICS Manager, Cancer Strategy and Development quality and data personnel, a representative metropolitan ICS data analyst, and the VCPMF project manager. Use of the information The aim of phase II of the project is to pilot the use of cancer quality and performance information to identify aspects of cancer care in need of further analysis, investigation and quality intervention. The dissemination of these results is a first step to developing the systematic reporting, analysis and use of cancer data on a state-wide basis. The data are intended to be used by each of the Integrated Cancer Services (ICS) for internal review and analysis in accordance with their own governance processes. The accompanying Guide to use of information is a companion document to these results. The guide provides for an overview of the project s framework for quality and performance improvement and charts the proposed flow of information, timelines and deliverables across organisations involved in the analysis of the cancer performance data. Cancer Performance Indicator Results 2013 Data 5

6 Data The Victorian Cancer Registry (VCR) The Victorian Cancer Registry (VCR) is a population-based cancer registry aiming to provide comprehensive, accurate and timely information for cancer control. Hospitals, pathology laboratories and cancer screening registers are required to notify the VCR of all cancers. These notifications are consolidated into individual tumour records. The VCR data collection started in 1982, and the subset of data that is linked includes any Victorian diagnosed between Although the most recent VCR data is 2014, the most recent year for which patient s addresses have been geocoded is As geocoding is needed to allocate each patient to an ICS of residence, this report includes only the subset of Victorians diagnosed in Administrative data Administrative datasets refer to those data collections made up of information routinely collected by departments and agencies, often mandated by statute, during the delivery of a service. These data are often collected for the purposes of reimbursement of activities, research, policy, planning, monitoring and evaluation (Allen, 2013). The datasets usually contain both identifiable information and summary content information (e.g. diagnosis, date and type of treatment) which relates to individuals. Data sets used for these indicators are: Victorian Admitted Episodes Dataset (VAED) provides a comprehensive dataset of the causes, effects and nature of illness, and the use of health services in Victoria. All Victorian public and private hospitals, including rehabilitation centres, extended care facilities and day procedure centres, report a minimum set of data for each admitted patient episode. Victorian Emergency Minimum Dataset (VEMD) comprises de-identified demographic, administrative and clinical data detailing presentations at designated emergency departments in Victorian public hospitals. Data linkage Data linkage is a method of bringing together information about people, places and events in a way that protects the privacy of individuals. It is used to link different health and health-related data collections for a range of population-based studies and performance monitoring and evaluation activities. Victorian Data Linkages (DHHS-VDL) provides population-wide linked data for research and quality improvement while adhering to Victorian privacy principles. Data linkage validation Linked data may create additional concerns about error if cases are not linked accurately. It is important that data linkage validation is undertaken wherever possible to identify factors able to compromise the quality of data. The validation of data linkages is an ongoing effort to accurately appraise the quality of results. Cancer Performance Indicator Results 2013 Data 6

7 General data statement Patients with multiple primary cancer diagnoses within 365 days of each other were excluded (n = 414), because of potential difficulties the assignment of treatment to each of the tumours. Data resources and support The VCPMF project would like to acknowledge the support and resources of the Department of Health and Human Services as custodians of the linked data. Tumour stream definitions Table 2: Tumour stream definitions by International Classification of Disease (ICD) code Tumour Stream Breast Central Nervous System (CNS) Colorectal Cancer (CRC) Endocrine Thyroid (ET) Genito Urinary (GU) Gynaecological (Gyn) Haematological (Haem) Head and Neck (HN) Lung Oesophago Gastric (OG) Skin Melanoma (Skin) Upper Gastrointestinal (UGI) Sub-tumour stream / cancer types Colon Ovarian ICD10 C50 C70, C71, C72 C18, C19, C20 C73, C74, C75 C60, C61, C62, C63, C64, C65, C66, C67, C68 C51, C52, C53, C54, C55, C56, C57, C58 C81, C82, C83, C84, C85, C86, C87, C88, C89, C90, C91, C92, C93, C94, C95, C96 C01, C02, C03, C04, C05, C06, C07, C08, C09, C10, C11, C12, C13, C14 C33, C34 C15, C16 C43 C15, C16, C17, C22, C23, C24, C25, C26 C18 C481, C482, C569, C570, C578, C579 Integrated Cancer Service (ICS) abbreviations Table 3: ICS abbreviations and diagram abbreviations Integrated Cancer Service (ICS) Barwon South Western Regional Integrated Cancer Service (BSWRICS) Gippsland Regional Integrated Cancer Service (GRICS) Grampians Regional Integrated Cancer Service (GICS) Hume Regional Integrated Cancer Service (Hume RICS) Loddon Mallee Integrated Cancer Service (LMICS) North Eastern Melbourne Integrated Cancer Service (NEMICS) Paediatric Integrated Cancer Service (PICS) Southern Melbourne Integrated Cancer Service (SMICS) Western & Central Melbourne Integrated Cancer Service (WCMICS) Diagram abbreviation BSW GR G HR LM NEM PICS SM WCM Cancer Performance Indicator Results 2013 Data 7

8 Cancer performance indicator results Performance indicator (PI) 1: Emergency presentation and new cancer diagnosis (2013) Summary specifications OCP Step(s): [2] Presentation, initial investigations and referral [3] Diagnosis and treatment planning Description / definition: Rationale: Denominator: Percentage of patients newly diagnosed with cancer during and following presentation to an emergency department (ED) Emergency department presentation is a surrogate for late diagnosis and has implications for access to health services. Even in countries with high performing health care systems up to one in four patients with cancer are diagnosed in emergency (NHS England, 2015). Emergency diagnosis is associated with poorer outcomes (One-year relative survival) for almost all cancer types, and may be particularly stressful for patients (Elliss-Brookes, 2012). Two key factors determining outcomes for many cancers are the route to diagnosis and stage at diagnosis. Number of people diagnosed with cancer who had presented to an ED within 4 weeks prior to diagnosis Numerator: i) For that episode of care (prior to separation) or within 4 weeks of date of presentation Tumour streams / cancer types: Breast, Central Nervous System, Colorectal, Endo-Thyroid, Genitourinary, Gynaecological, Haematological, Head & Neck, Lung, Skin and Upper Gastrointestinal Data collection statement For each of the newly diagnosed cancer patients in 2013, a patient was flagged as having an ED presentation if the arrival date at the ED (in VEMD) fell within 28 days prior to their diagnosis date (in VCR; including the date of diagnosis). The indicator takes into account any presentation to a designated emergency department, irrespective of its relation to the cancer diagnosis. Emergency presentation for patients that present to a hospital without a designated emergency department are currently not included in this indicator. Data sources VCR, VEMDS Cancer Performance Indicator Results 2013 Data 8

9 Overall Victoria PI 1: Emergency presentation and new cancer diagnosis (2013) Figure 1: Emergency presentation and new cancer diagnosis (PI - 1) for all tumour streams over all Victoria in calendar year 2013 with associated data table. Comment The number of cases with ED presentation is displayed on top of the blue bars and the total number of diagnoses in 2013 under the blue bars. For each of the newly diagnosed cancer patients in 2013, a patient was flagged as having an emergency department (ED) presentation if the arrival date at the ED (in VEMD) fell within 28 days of the diagnosis date (in VCR; including the date of diagnosis). It includes any presentation to the ED regardless of its relation to cancer diagnosis (i.e., if you were admitted to ED because of something totally different, but were picked up as a new cancer diagnosis). Overall,17.5% of Victorians diagnosed with a primary cancer had emergency department (ED) presentation in the 28 days prior their diagnosis. Marked variation exists among tumour streams, ranging from with 61.6% (CNS) to 3.5% (Skin). Because not all presentations to an ED result in its reporting to the Victorian Emergency Minimum Dataset (VEMDS), the indicator will result in an underestimation in general, and potentially more so in rural areas. Cancer Performance Indicator Results 2013 Data 9

10 Integrated Cancer Service Figures 2-4 display the same 2013 data as in Figure 1 but stratified by patient ICS of residence. Figure 2: Emergency presentation and new cancer diagnosis (PI - 1) for breast, CNS, CRC and endocrine-thyroid tumour streams for each patient ICS of residence in calendar year 2013 (the orange line is the Victorian average). Comment Please note the variation in scale on the vertical axis for each of the tumour streams. Cancer Performance Indicator Results 2013 Data 10

11 Figure 3: Emergency presentation and new cancer diagnosis (PI - 1) for genito-urinary, gynaecological, haematological, and head and neck tumour streams for each patient ICS of residence in calendar year 2013 (the orange line is the Victorian average). Comment Please note the variation in scale on the vertical axis for each of the tumour streams. Cancer Performance Indicator Results 2013 Data 11

12 Figure 4: Emergency presentation and new cancer diagnosis (PI - 1) for lung, skin and upper gastrointestinal tumour streams for each patient ICS of residence in calendar year 2013 (the orange line is the Victorian average). * Data is statistically significant at the (0.625%) level (.05 divided by 8 comparisons for the 8 ICS included). That is, the P value of is when each ICS is compared to the state average, less the ICS concerned. It is a very stringent limit which aims to limit false positives. Comment Please note the variation in scale on the vertical axis for each of the tumour streams. Regional variation might be associated with variation in access to EDs. WCMICS for upper gastrointestinal was statistically significant at p < That is, there are proportionally more patients presenting via EDs in WCMICS at this limit. This is the only statistically significant result for all of the PI 1 results across tumour streams. Cancer Performance Indicator Results 2013 Data 12

13 Performance indicator (PI) 6: Stage at diagnosis (2013) Summary specifications OCP Step(s): Description / definition: [3] Diagnosis and treatment planning i) The proportion of new cancer diagnoses missing a Victorian Cancer Registry (VCR) derived stage ii) VCR derived stage at diagnosis Rationale: Cancer stage at diagnosis is critical in determining a prognosis and is important for making treatment decisions (Brierley, 2013). Information on stage is also used for planning and managing cancer services, and for evaluating, measuring and reporting on cancer treatment patterns and outcomes (CQCO, 2016). Identifying stage at diagnosis from the linked Victorian Cancer Registry Victorian Admitted Episode Dataset is a recommendation from the 2014 Victorian Colorectal Cancer Summit (VICS, 2015). From 1 July 2013, the revised Cancer (Reporting) Regulations 2012 mandate the reporting of cancer stage or extent of disease at diagnosis, when known (State of Victoria, 2012). Denominator: i) Total number of new cancer diagnoses ii) Total number of new cancer diagnoses with non-missing stage at diagnosis Numerator: i) Number of diagnoses with missing VCR derived stage ii) VCR derived stage at diagnosis Tumour streams / cancer types: Breast, Colorectal Data collection statement Victorian Cancer Registry derived stage (or RD-Stage, see below) is currently only available for breast and colorectal cancer for this data linkage. Exclusions include: non-stageable incident cancer cases, non-incident cases, staging after neo-adjuvant therapy and patients diagnosed at death. Non-missing stage is the number of diagnoses with a recorded stage derived by the VCR. Missing stage at diagnosis is the number of diagnoses recorded by the VCR with a derived stage missing. VCR- or registry-derived stage (RD Stage) There are multiple classification systems of cancer staging. The VCR derives stage at diagnosis by applying the American Joint Committee on Cancer (AJCC) TNM staging principles. The information to inform the specific T, N and M components for staging is a best estimate derived from summary sources at time of diagnosis. Hence, a combination of clinical and pathological data is used for deriving stage according to established business rules. Occasionally, hospital notifications are the only source of information regarding distant metastases, which may result in an underreporting of stage IV. Preliminary validation work undertaken by the VCR has demonstrated that the staging dataset is suitable for use in population analysis (Thursfield, 2016). Data sources VCR Cancer Performance Indicator Results 2013 Data 13

14 Overall Victoria PI 6: VCR derived summary stage at diagnosis (2013) Figure 5: Recorded stage at cancer diagnosis (PI - 6) for breast and colorectal cancer for all Victoria in calendar year Comment Figure 5 displays the percentage of each stage, derived by VCR, for those diagnoses of breast and colorectal cancer that had recorded stage in calendar year Above the bars is displayed the total number of diagnoses with a recorded stage that was derived by VCR. For example, in 2013 there was a total of 3,767 breast cancer diagnoses that had a recorded stage derived by VCR of which approximately 43% were deemed as stage I, 41% as stage II and so on. Below the bars is displayed the number of diagnoses that were missing recorded stage and its percentage of the total number where total in this instance means the number with and without recorded staging. For example 291 is 7% of ,767. The 2013 data will become more meaningful when it is able to be compared with other data from other years (e.g and 2015 data linkages). The intention is to extend the number of tumour streams and cancer types for comparison as the capability develops. Cancer Performance Indicator Results 2013 Data 14

15 Integrated Cancer Services Figures 6 and 7 display the same 2013 data as in Figure 5 but stratified by ICS of patient residence. The horizontal lines in Figures 6 and 7 pertain to the overall Victorian percentage for each stage category signified by the colour of the line. The variation was assessed statistically, but none were significant at p = Similar to Figure 5, below the bars is displayed the number of diagnoses that were missing recorded stage and its percentage in total with the missing included with the number of cases. Please note miss. is an abbreviation for missing. Figure 6: Recorded stage at cancer diagnosis (PI - 6) for breast cancer for each patient ICS of residence in calendar year The horizontal lines pertain to the Victorian overall percentage for each stage and are coloured accordingly. Cancer Performance Indicator Results 2013 Data 15

16 Figure 7: Recorded stage at cancer diagnosis (PI - 6) for colorectal cancer for each patient ICS of residence in calendar year The horizontal lines pertain to the Victorian overall percentage for each stage and are coloured accordingly. Cancer Performance Indicator Results 2013 Data 16

17 Performance indicator (PI) 8: Reporting ECOG performance status (2013) Summary specifications OCP Step(s): Description / definition: [3] Diagnosis Proportion of patients with documented Eastern Cooperative Oncology Group (ECOG) performance status i) ECOG performance status recorded within 30 days of diagnosis Rationale: ECOG Performance status is one of the most important predictors of cancer outcomes and a key determinant of treatment. (Davidoff, 2010; Oken, 1982). Changes to the Cancer (Reporting) Regulations 2012 stipulate the reporting of ECOG performance status score (if known) at the time of diagnosis to the Victorian Cancer Registry. Denominator: i) Total number of new cancer diagnoses Numerator: i) Number of new diagnoses with ECOG performance status recorded within 30 days of diagnosis ii) ECOG performance status Tumour streams / cancer types: Breast, Central Nervous System, Colorectal, Endocrine & Thyroid, Genito-urinary, Gynaecological, Haematological, Head & Neck, Lung, Skin, and Upper Gastrointestinal Data collection statement ECOG status has only been required for collection by the VCR in the last few years. The 2013 linked VCR dataset includes ECOG as a variable, however, it seems to be 9 (Unknown) for all patients. Data sources VCR Cancer Performance Indicator Results 2013 Data 17

18 Overall Victoria PI 8: Reporting ECOG performance status (2013) Figure 8: ECOG performance status recorded within 30 days of diagnosis. Comment ECOG status for all patients diagnosed in 2013 is recorded as unknown. As the VCR has recently started to collect this data in a more systematic way, it is anticipated that results for the 2015 calendar year will be more complete. Integrated Cancer Services There is no available breakdown by ICS for ECOG performance status. Cancer Performance Indicator Results 2013 Data 18

19 Performance indicator (PI) 13: Timeliness of initial treatment after cancer diagnosis (2013) Summary specifications OCP Step(s): [3] Diagnosis and treatment planning [4] Treatment Description / definition: Rationale: Denominator: Timeliness of initial cancer treatment from date of diagnosis to date of initial treatment Waiting times are clinically relevant and should be informed by evidencebased guidelines where they exist (CCV, 2015). The OCPs for cancer care recognise that shorter timeframes for appropriate consultations and treatment can reduce patient distress (DHHS, 2015). Wait times can serve as a gauge of how well the cancer system is working (Evans, 2016). Total number of patients with new cancer diagnoses who commence initial treatment within 6 months Numerator: i) Proportion of patients with time (days) from date of diagnosis to day one (1) of chemotherapy or surgery (whichever occurred first) within 4 weeks ii) Proportion of patients with time (days) from date of diagnosis to day one (1) of chemotherapy within 4 weeks (where chemotherapy is the 1st treatment) iii) Proportion of patients with time (days) from date of diagnosis to date of surgery within 4 weeks (if surgery is the 1st treatment) Tumour streams / cancer types: Breast, CNS, Colorectal, Endocrine-Thyroid, Genito-Urinary, Gynaecological, Haematological, Head & Neck, Lung, Oesophago- Gastric and Skin Data collection statement The date of diagnosis is taken as a proxy for initial referral, decision to treat or ready for care as recommended by the OCPs across specific tumour streams. The intent of treatment is not captured. Neo-adjuvant therapy is considered as a first treatment. The data does not capture oral chemotherapy. A time interval of four weeks is utilised across tumour streams and cancer types notwithstanding small variations in OCP recommendations. For example, non-small cell lung cancer (NSCLC) chemotherapy within three weeks of the ready for care date in lung cancer. Currently radiotherapy is not included. This will change with a new 2015 data linkage. No risk-adjustment has been made. Data sources VCR, VAED Cancer Performance Indicator Results 2013 Data 19

20 Overall Victoria PI 13: Timeliness of initial treatment after cancer diagnosis (2013) Figure 9: Funnel plot for the proportion of patients that received surgery or chemotherapy within 4 weeks of diagnosis (PI - 13) for all tumour streams in Comment Funnel plots show an average rather than a time series. A point in the funnel plot which lies on the Victorian average (solid horizontal black line) could still include periods of good and bad runs. Services with results that structurally differ from the Victorian average will be picked up irrespective of their volume. Cancer Performance Indicator Results 2013 Data 20

21 For all of the funnel plots the inner and outer lines represent statistical significance at the 0.05 (5%) and (0.1%) levels respectively. See also the accompanying document Guide to use of information for further advice on how to read a funnel plot. Over all tumour streams, 69% of patients started their initial treatment (chemotherapy or surgery) within 4 weeks of diagnosis. For all tumour streams, where chemotherapy was the initial treatment, 56% commenced within 4 weeks of diagnosis. Where surgery was the initial treatment, 75% commenced within 4 weeks of diagnosis. Variation between tumour streams exist, ranging from 56.9% (GU) to 94.8% (CNS) irrespective of initial treatment. Among the tumour streams endocrine-thyroid, genito-urinary, head and neck, oesophago-gastric and skin cancers were much less likely to have had initial treatment within 6 months of diagnosis. The variation for these tumour streams is statistically highly significant (p < 0.001). Cancer Performance Indicator Results 2013 Data 21

22 Integrated Cancer Services Figures 10 to 21 display the same data as in figure 9 but stratified by ICS of patient residence. The graph for all tumour streams is presented in Figure 10. This is followed by each tumour stream separately (Figures 11-21). Each tumour stream has been stratified for all patients that received chemotherapy or surgery, those that had chemotherapy as their first treatment and those that had surgery as their first treatment. Where the numbers for chemotherapy as first treatment are less than 10, a single graph for surgery as first treatment has been provided (e.g. CNS, ET & skin). Figure 10: The proportion of patients within their ICS of residence that received surgery or chemotherapy within 4 weeks of diagnosis (PI - 13) for all tumour streams. Cancer Performance Indicator Results 2013 Data 22

23 Figure 11: The proportion of patients within their ICS of residence that received surgery or chemotherapy within 4 weeks of diagnosis (PI - 13) for breast cancer. Cancer Performance Indicator Results 2013 Data 23

24 Figure 12: The proportion of patients within their ICS of residence that received surgery within 4 weeks of diagnosis (PI - 13) for CNS cancer. Comment Only the results for surgery as first treatment for CNS within 4 weeks of diagnosis is shown as the number of patients receiving chemotherapy as first treatment is under 10. Cancer Performance Indicator Results 2013 Data 24

25 Figure 13: The proportion of patients within their ICS of residence that received surgery or chemotherapy within 4 weeks of diagnosis (PI - 13) for colorectal cancer. Cancer Performance Indicator Results 2013 Data 25

26 Figure 14: The proportion of patients within their ICS of residence that received surgery within 4 weeks of diagnosis (PI - 13) for endocrine-thyroid cancer. Comment Only the results for surgery as first treatment for endocrine-thyroid cancer within 4 weeks of diagnosis is shown as the number of patients receiving chemotherapy as first treatment is under 10. Cancer Performance Indicator Results 2013 Data 26

27 Figure 15: The proportion of patients within their ICS of residence that received surgery or chemotherapy within 4 weeks of diagnosis (PI - 13) for genito-urinary cancer. Cancer Performance Indicator Results 2013 Data 27

28 Figure 16: The proportion of patients within their ICS of residence that received surgery or chemotherapy within 4 weeks of diagnosis (PI - 13) for gynaecological cancer. Cancer Performance Indicator Results 2013 Data 28

29 Figure 17: The proportion of patients within their ICS of residence that received surgery or chemotherapy within 4 weeks of diagnosis (PI - 13) for haematological cancer. Cancer Performance Indicator Results 2013 Data 29

30 Figure 18: The proportion of patients within their ICS of residence that received surgery or chemotherapy within 4 weeks of diagnosis (PI - 13) for head and neck cancer. Cancer Performance Indicator Results 2013 Data 30

31 Figure 19: The proportion of patients within their ICS of residence that received surgery or chemotherapy within 4 weeks of diagnosis (PI - 13) for lung cancer. Cancer Performance Indicator Results 2013 Data 31

32 Figure 20: The proportion of patients within their ICS of residence that received surgery or chemotherapy within 4 weeks of diagnosis (PI - 13) for oesophago-gastric cancer. Cancer Performance Indicator Results 2013 Data 32

33 Figure 21: The proportion of patients within their ICS of residence that received surgery within 4 weeks of diagnosis (PI - 13) for skin cancer. Comment Only the results for surgery as first treatment for skin cancer within 4 weeks of diagnosis is shown as the number of patients receiving chemotherapy as first treatment is under 10. Cancer Performance Indicator Results 2013 Data 33

34 Performance indicator (PI) 16: Timeliness of adjuvant chemotherapy for stage III colon cancer (2013) Summary specifications OCP Step(s): Description / definition: Rationale: Denominator: [4] Treatment The proportion of patients who did receive adjuvant chemotherapy intravenously within 8 weeks of surgery for stage III colon cancer Adjuvant chemotherapy improves outcomes for patients with colon cancer (Bayraktar, 2010). Waiting times are clinically relevant and also significant from a patient perspective. The Guidelines for timely initiation of chemotherapy (DHHS, 2015) and the OCP for people with colorectal cancer both recommend adjuvant chemotherapy should commence as soon as the patient is medically fit and within 8 weeks of the date of surgery (OCP, 2015) [Evidence level-grade III-C]. It is a recommendation of the 2014 Victorian Colorectal Cancer Summit to report on this (VICS, 2015). Total number of patients diagnosed with stage III colon cancer who had surgery within 90 days of the diagnosis date as reported in the VCR, and had started intravenous chemotherapy within 182 days of their surgery. Numerator: i) Number of patients who received adjuvant chemotherapy intravenously within 8 weeks of surgery Tumour streams / cancer types: Colon cancer Data collection statement The date of diagnosis is taken as a proxy for the ready for care date recommended in the Guidelines for timely initiation of chemotherapy (DHHS, 2015). Data sources VCR, VAED Cancer Performance Indicator Results 2013 Data 34

35 Integrated Cancer Services PI 16: Timeliness of adjuvant chemotherapy for stage III colon cancer (2013) Figure 22: Proportion of patients who did receive adjuvant chemotherapy intravenously within 8 weeks of surgery (PI 16) for stage III colon cancer by ICS of patient residence. Comment This plot included only patients with CRC and derived stage III at diagnosis. Of the patients with chemotherapy following surgery, 40% received their first chemotherapy round within 8 weeks of surgery. This state average of 40% is noteworthy and well below the goals outlined in the OCP for colorectal cancer. The percentage of patients within their ICS of residence varied between 28.6% and 47.1% for BSWRICS and SMICS respectively. As all the ICS appear within 95% control limits, there were no statistical differences at the 5% level (p = 0.05) when the ICS are compared to the overall average of 40%. Cancer Performance Indicator Results 2013 Data 35

36 Performance indicator (PI) 17: Number of lymph nodes examined during colon cancer surgery (2013) Summary specifications OCP Step(s): Description / definition: Rationale: Denominator: [4] Treatment The proportion of patients with colon cancer (Stage I, II or III) who undergo surgical resection where 12 or more lymph nodes are pathologically examined Examining sufficient lymph nodes at the time of surgery assists in the assessment of the extent of cancer and the appropriate use of adjuvant therapy for node-positive patients (Chang, 2007; Wong, 2009). Maximising the number of lymph nodes resected and analysed enables reliable staging which influences decision making [NHS Quality Improvement Scotland, 2015: Evidence level Grade: III-2]. Routine reporting (by region and health service) of the number of nodes examined during major colorectal cancer surgery is a recommendation of the 2014 Victorian Colorectal Cancer Summit (VICS, 2015). Total number of patients diagnosed with stage I,II or III colon cancer who undergo surgical resection Numerator: i) The number of patients with 12 or more lymph nodes removed and examined Tumour streams / cancer types: Colon cancer Data collection statement The number of lymph nodes sampled at surgery is collected in the VCR, not the VAED. The date of surgery in which lymph nodes were collected is known at the VCR, however, it is not available in the current linked-vcr/vaed dataset. As a result, the lymph node data cannot be associated with specific surgical resections at this stage. The date of surgery will be available at a later stage. Via the surgery dates, the matching admission could be found and as a result we could 1) provide insight into the proportion of surgeries in which the number of lymph nodes was not reported to the VCR (part ii of the indicator) and 2) provide data at a health service level. Stage IV colon cancer and those patients unstaged have been excluded. Patients who undergo transanal endoscopic microsurgery or transanal resection of tumour. Data sources VCR Cancer Performance Indicator Results 2013 Data 36

37 Integrated Cancer Services PI 17: Number of lymph nodes examined during colon cancer surgery for stages I III (2013) Figure 23: The proportion of colon cancer surgeries for patients with VCR derived stage between I III that had 12 or more lymph nodes examined, within patient ICS of residence. Comment Overall, 82% of stage I, II, or III colon cancer patients with a reported number of lymph nodes taken had 12 or more lymph nodes taken. The percentage within ICS of residence varied between 68.6% to 89.4% for Hume RICS and BSWRICS respectively. When compared to the overall average, BSWRICS was outside the 95% control limits and so was statistically significant at the 5% level (p = 0.05). Hume RICS is very close to the 99.9% control limit and so compared to the overall average was statistically significant at approximately the 0.1% level (p = 0.001). Hume had a lower percentage of approximately 68% and Barwon a higher percentage of approximately 89%. These differences were statistically significant at approximately p < 0.01 and p < 0.05 respectively. This data is available in numeric form upon request for other cohorts of numbers of lymph nodes derived. Cancer Performance Indicator Results 2013 Data 37

38 Glossary Adjuvant chemotherapy Intravenous chemotherapy given after surgery. Chemotherapy start date The date of commencement of the first day of the first cycle of (adjuvant) intravenous chemotherapy given to a patient who is receiving care for cancer. Date of diagnosis The earliest date of evidence of cancer according to site and morphology. In most cases the date of diagnosis would be the date of biopsy or date of surgery which led to the specimen in which cancer was detected. Date of surgery The date of admission of the hospitalisation in which cancer related surgery was performed on a patient. Initial cancer treatment where intravenous chemotherapy, surgery, is the first treatment. In this report, radiotherapy was excluded. Optimal cancer care pathway the key principles and practices required at each stage of the care pathway to guide the delivery of consistent, safe, high-quality and evidence-based care. Performance status A measure of how well a patient is able to perform ordinary tasks and carry out day-to-day activities.(0=fully active 4=Completely disabled, 9=Unknown) Registry-derived stage (RD-Stage) Registry-derived stage (RD-Stage) is defined as the best estimate of summary TNM stage at time of diagnosis as derived by cancer registries (e.g. Victorian Cancer Registry) from available data sources for use in population data analysis. Cancer Performance Indicator Results 2013 Data 38

39 References Allen J, Holman CD, Meslin EM, Stanley F Privacy protectionism and health information: is there any redress for harms to health? Journal of Law and Medicine: 21(2); Bayraktar S, Bayraktar UD, Rocha-Lima CM Timing of adjuvant and neoadjuvant therapy in colorectal cancers. Clinical Colorectal Cancer; 9: Brierley JD, et al The value of collecting population-based cancer stage data to support decisionmaking at organisational, regional and population levels. Healthcare Quarterly: 16(3): Cancer Council Victoria Optimal Cancer Pathways. Accessed on 26 October Thursfield V, Farrugia H. Cancer in Victoria: Statistics & Trends Cancer Council Victoria, Melbourne 2016: p. 25. Cancer Quality Council of Ontario (CQCO). Reporting of Cancer Stage at Diagnosis. Accessed on 26 October 2016, at: Chang GJ, Rodriquez-Bigas MA, Skibber JM, Moyer VA Lymph node evaluation and survival after curative resection of colon cancer: systematic review. Journal of the National Cancer Institute: 99(6); Davidoff AJ, et al Development of a performance status prediction model for use in administrative data analyses. Journal of Clinical Oncology 28:15s (Suppl;abstr 6006). Department of Health and Human Services Guidelines for timely initiation of chemotherapy. A proposed framework for access to medical oncology and haematology cancer clinics and chemotherapy services in Victoria. Elliss-Brookes, L et. al Routes to diagnosis for cancer determining the patient journey using multiple routine data sets. British Journal of Cancer: 107(8): Evans SM, Earnest A, Bower W, Senthuren M, McGlauglin P, Stirling R Timeliness of lung cancer care in Victoria: a retrospective cohort study. The Medical Journal of Australia: 204(2); 75 NHS England and Public Health England Progress in improving cancer services and outcomes in England. A report by the Comptroller and Auditor General. National Audit Office: NHS Scotland. Scottish Cancer Taskforce. (2015 v2.1) Colorectal Cancer Clinical Quality Performance Indicators. Oken M, Creech R, Tormey D, et al Toxicity and response criteria of the Eastern Cooperative Oncology Group. American Journal of Clinical Oncology.: 5; State of Victoria. Cancer (Reporting) Regulations S.R. No. No8/2012. Accessed on 26 October Victorian Integrated Cancer Services (VICS) Victorian Colorectal Cancer Summit Newsletter (June 2015). Wong SL Lymph node counts and survival rates after resection for colon and rectal cancer. Gastrointestinal Cancer Research: 3(2 Suppl 1; S33-S35) Cancer Performance Indicator Results 2013 Data 39

40 Appendices Appendix 1: Indicator selection and development flow chart Appendix 2: Integrated Cancer Service by health service data Cancer Performance Indicator Results 2013 Data 40

41 APPENDIX 1 - INDICATOR SELECTION AND DEVELOPMENT FLOW CHART M a y 2016 A flow chart of the indicator selection and development process (Phase I) 1

42 APPENDIX 2 GICS BY HEALTH SERVICE VCPMF PHASE II PILOT D e c e m b e r 2016 Appendix 2 Grampians Regional Integrated Cancer Service (GICS) Integrated Cancer Service by health service (campus) The results for Performance Indicator 13 have been chosen by the VCPMF Steering Group for display by campus. The following graphs show the proportion of patients within each Victorian health service that received surgery or chemotherapy within 4 weeks of diagnosis (PI 13). The represented data by campus is only for colorectal cancer and lung cancer. For each Integrated Cancer Service there are three funnel plots presented for each of the represented tumour streams. 1. The proportion of patients that received chemotherapy or surgery as a first treatment within 4 weeks of diagnosis 2. The proportion of patients that had chemotherapy as their first treatment within 4 weeks of diagnosis 3. The proportion of patients that had surgery as their first treatment within 4 weeks of diagnosis Each orange lined circle within the funnel plot represents a health service across the state. Each ICS will receive the same graph, except that the version received by an ICS will have their respective health services block coloured with the campus number identified. A particular ICS will only see their health services coloured and numbered. The health services in other ICSs are differentiated by the orange lined circle but are not identifiable. Where there is no applicable health service this is marked. The orange lined circles representing the campuses across the state are still shown. There are cases where two or more campuses are displayed in close proximity on the funnel plot and the campus numbers are obscured. These can be identified by a process of elimination. Further information can be obtained upon request. Cancer Performance Indicator Results 2013 Data 1

43 APPENDIX 2 GICS BY HEALTH SERVICE VCPMF PHASE II PILOT D e c e m b e r 2016 Grampians Regional Integrated Cancer Service (GICS) Figure 24: The proportion of patients within their ICS of residence by campus that received surgery or chemotherapy within 4 weeks of diagnosis (PI - 13) for colorectal cancer (2013). Comment: The GICS campuses are coloured in purple with specific campuses numbered. Each orange lined circle within the funnel plot represents a de-identified health service in an other ICS. Cancer Performance Indicator Results 2013 Data 2

44 APPENDIX 2 GICS BY HEALTH SERVICE VCPMF PHASE II PILOT D e c e m b e r 2016 Figure 25: The proportion of patients within their ICS of residence by campus that received chemotherapy within 4 weeks of diagnosis (PI - 13) for colorectal cancer (2013). Cancer Performance Indicator Results 2013 Data 3

45 APPENDIX 2 GICS BY HEALTH SERVICE VCPMF PHASE II PILOT D e c e m b e r 2016 Figure 26: The proportion of patients within their ICS of residence by campus that received surgery within 4 weeks of diagnosis (PI - 13) for colorectal cancer (2013). Cancer Performance Indicator Results 2013 Data 4

46 APPENDIX 2 GICS BY HEALTH SERVICE VCPMF PHASE II PILOT D e c e m b e r 2016 Figure 27: The proportion of patients within their ICS of residence by campus that received surgery or chemotherapy within 4 weeks of diagnosis (PI - 13) for lung cancer (2013). Comment: The GICS campuses are coloured in purple with specific campuses numbered. Each orange lined circle within the funnel plot represents a de-identified health service in an other ICS. Cancer Performance Indicator Results 2013 Data 5

47 APPENDIX 2 GICS BY HEALTH SERVICE VCPMF PHASE II PILOT D e c e m b e r 2016 Figure 28: The proportion of patients within their ICS of residence by campus that received chemotherapy within 4 weeks of diagnosis (PI - 13) for lung cancer (2013). Cancer Performance Indicator Results 2013 Data 6

48 APPENDIX 2 GICS BY HEALTH SERVICE VCPMF PHASE II PILOT D e c e m b e r 2016 Figure 29: The proportion of patients within their ICS of residence by campus that received surgery within 4 weeks of diagnosis (PI - 13) for lung cancer (2013). Cancer Performance Indicator Results 2013 Data 7