ORIGINAL ARTICLE
The Clinical Respiratory Journal
What is the minimum number of patients for quality control of lung cancer management in Norway? Knut Skaug1,2, Geir E. Eide3,4 and Amund Gulsvik2 1 Department of Medicine, NO-5504 Haugesund Hospital, Health Region of Fonna, Haugesund, Norway 2 Section of Thoracic Medicine, Department of Clinical Science, University of Bergen, Bergen, Norway 3 Centre for Clinical Research, Haukeland University Hospital, Bergen, Norway 4 Research Group for Lifestyle Epidemiology, Department of Global Public Health and Primary Care, University of Bergen, Bergen, Norway
Abstract Introduction: There are few data available on the optimal number of lung cancer patients needed to generate and compare estimates of quality between units managing lung cancer. The number of lung cancer patients per management unit varies considerably in Norway, where there are 42 hospitals that treated between 1 and 454 lung cancer patients in 2011. Aims: To estimate the differences in quality indicators that are of sufficient importance to change a pulmonary physician’s lung cancer management program, and to estimate the size of the patient samples necessary to detect such differences. Method: Twenty-six physicians were asked about the relative differences from a national average of quality indicators that would change their own lung cancer management program. Sample sizes were calculated to give valid estimates of quality of a management unit based on prevalence of quality indicators and minimally important differences (MID). Results: The average MID in quality indicators that would cause a change in management varied from 18% to 24% among 26 chest physicians, depending on the indicator. Conclusions: To generate precise estimates for quality control of lung cancer care in Norway, the number of management units must be reduced. Given the present willingness of chest physicians to change their procedures for management of lung cancer according to the results of quality control indicators, we recommend a maximum of 10 units with a minimum of 200 incident lung cancer patients per year for each management center. Please cite this paper as: Skaug K, Eide GE and Gulsvik A. What is the minimum number of patients for quality control of lung cancer management in Norway? Clin Respir J 2015; ••: ••–••. DOI:10.1111/crj.12274.
Introduction The quality of care in lung cancer is of major concern for patients and health-care providers (1). The purpose of national guidelines (2) and a quality registry is to improve the care of these patients (1). Few data are available on how national or regional management programs for lung cancer care are followed. The national cancer registries of the Nordic countries are
The Clinical Respiratory Journal (2015) • ISSN 1752-6981 © 2015 John Wiley & Sons Ltd
Key words lung cancer – indicators – patient volume – quality Correspondence Knut Skaug, MD, PhD, Department of Medicine, NO-5504 Haugesund Hospital, Health Region of Fonna, P.O. Box 2170, NO-5104 Haugesund, Norway. Tel: +47 5 273 20 00 Fax: +47 5 273 20 02 email: [email protected] Received: 27 March 2014 Revision requested: 17 October 2014 Accepted: 20 January 2015 DOI:10.1111/crj.12274 Authorship and contributorship Knut Skaug: planning of the project; collecting and analyzing the data; writing the manuscript. Geir E. Eide: analyzing the data; contributed to designing the study and manuscript. Amund Gulsvik: planning and designing of the project; writing the manuscript; senior author and adviser. Conflicts of interest The authors have stated explicitly that there are no conflicts of interest in connection with this article.
known to be of good quality (3, 4). Indicators used to compare, coordinate and improve lung cancer care have been established in England (5), Germany (6) and Denmark (7). Regional differences in survival within a country have been documented in Spain (8), Sweden (9) and Norway (10). However, little is known about how many patients a lung cancer management unit should treat per year to give meaningful estimates of quality (11). Quality
1
Quality, lung cancer, indicators, patient volume
Skaug et al.
indicators of lung cancer management that have been used are the proportion of patients staged through multimodal investigation (12), diagnosis verified by histology or cytology (13), proportion of patients offered surgery (14), and the percentage of patients with 1-year survival (15). The aims of our study were to estimate the differences in quality indicators that are of sufficient importance to change a pulmonary physician’s lung cancer management program, and to estimate the size of the patient samples necessary to detect such differences.
Materials and methods Organization of lung cancer care in Norway Since 2002, Norway has been divided into four health regions, North, Middle, West and South-East, which are given their yearly performance targets by the governmental Department of Health. The annual numbers of incident cases of lung cancer in 2011 in the four health regions were 291, 370, 508 and 1551, respectively (15). However, the autonomy of management within the health regions is considerable for hospitals, thoracic surgeons, oncologists and chest physicians. In 2011, treatment for lung cancer was provided by 42 of the 59 hospitals in Norway (Fig. 1). Altogether, 33 hospitals reported the time from acceptance of a referred patient to the first tumor-directed treatment, and the number of patients per management unit varied from 1 to 454 patients of a total of 1618 patients. Fourteen hospitals treated fewer than 10 patients with lung cancer per year (16). Resection surgery and/or radiation treatment were performed in 10 hospitals.
National cancer registry The population-based Cancer Registry of Norway (CRN) receives clinical and pathology reports for all
cancer patients in the country. Death certificates are an additional source of data. The CRN has been shown to include more than 98% of all cancer patients in the country (3).
Lung cancer quality indicators We used the same measures of diagnostic and treatment quality indicators as were used for the survey of health care for lung cancer in Haugalandet, Norway (17). To obtain the national values for these quality indicators, we got from CRN the number of all incident lung cancer patients in Norway, as described in Table 1. Then we got in percent (i) the proportion of the patients who were staged (2), (ii) the proportion with a histological verification of the diagnosis (13), (iii) the proportion of patients with surgical resections (14) and (iv) 1-year survival (15).
Minimally important differences (MID) in quality indicators The method for estimating MID has been described in several previous studies (18, 19). Twenty-six physicians at the Department of Thoracic Medicine, Haukeland University Hospital, Bergen, completed a questionnaire, giving the percentage difference from the national average for each of the quality indicators that would lead to changes in their own management of lung cancer. They were informed of the national average of the four quality indicators (Table 1). Each physician was asked to tick a box for predetermined relative differences (5%, 10%, 20%, 30%, 40%, 50%, 75% or 100%) from the national average that would lead them to change their own management program of procedures to diagnose and treat lung cancer. Fourteen of the chest physicians had
Table 1. Number of patients (sample size) required to detect a 20% relative deterioration from the observed national average for no histology*, no staging,† surgical resection† and 1-year survival† and years to collect sufficient patients in management units
Indicators of care
Observed national average (%)
20% relative deterioration
Sample size (number of patients)
No histology No staging Surgical resection One-year survival
23.9 5.0 17.9 38.9
28.7% 6.0% 14.3% 31.1%
531 3283 686 242
Health regions years to collect sample
Center
South-East (1551)‡
West (508)‡
Middle (370)‡
North (291)‡
(200)§
0.3 2.1 0.4 0.2
1.0 6.5 1.4 0.5
1.4 8.9 1.9 0.7
1.8 11.3 2.4 0.8
2.7 16.4 3.4 1.2
*Reported from Cancer Registry of Norway (CRN) on 2720 incident lung cancer patients in 2011 in Norway and the four health regions. †Reported from CRN on 7741 incident lung cancer patients 2008–2010. ‡Incident lung cancer patients in the four health regions in Norway in 2011. §Incident lung cancer patients to be given accreditation as lung cancer management center.
2
The Clinical Respiratory Journal (2015) • ISSN 1752-6981 © 2015 John Wiley & Sons Ltd
Skaug et al.
Quality, lung cancer, indicators, patient volume
Figure 1. Location of hospitals in Norway taking part in lung cancer care (2011).
The Clinical Respiratory Journal (2015) • ISSN 1752-6981 © 2015 John Wiley & Sons Ltd
3
Quality, lung cancer, indicators, patient volume
Skaug et al.
No histology (%)
10
10
20
20
30
30
40
40
50
50
No staging (%)
0
5
10
15
20
25
0
Mean 22.5 (line), 95% CI (16.3, 28.7) (dotted)
5
10
15
20
25
Mean 23.8 (line), 95% CI (16.8, 29.4) (dotted)
One-year survival (%)
0
5
10
15
20
25
Mean 18.9 (line), 95% CI (14.1, 23.6) (dotted)
40 30 20 10
10
20
30
40
One-year survival (%)
50
50
Surgical treatment (%)
0
5
10
15
20
25
Mean 17.9 (line), 95% CI (13.0, 22.8) (dotted)
Figure 2. Relative percentages deviation (minimally important difference) from the national average on four quality indicators sufficient for the physician to change their management of lung cancer. Answers were given by 26 physicians at the Department of Thoracic Medicine, University of Bergen. The x axis denotes physicians by number 1–26. CI, confidence interval.
4 years or more of training in a chest clinic, while 12 had less than 4 years of training. Using Cytel Studio (version 10.0; Cytel Corporation, Cambridge, MA, USA) and one-sided tests, we calculated the sample sizes required to obtain a preset deviation of 20% from the national average for the four quality indicators, with significance set at P < 0.05 and power at 0.80 and 0.90.
Results Deviation from the national average causing change of management The deviation in the proportion of patients with unknown stage that would induce change by the physicians in their lung cancer management was on average 23% and for unknown histology 24% (see Fig. 2). The relevant deviations in the percentage of
4
patients undergoing surgery and in 1-year survival were on average 19% and 18%, respectively. In a subanalysis comparing these results between physicians with 3 years of experience in thoracic medicine or less with those who had 4 years or more, there were no significant differences.
Necessary number of samples to detect a 20% deviation from national average The sample sizes required to detect a deviation of 20% for the four quality indicators compared with the national average with a power of 80% at a significance level of P ≤ 0.05 are shown in Table 1. The sample sizes varied from 242 patients for 1-year survival to 3283 patients for lack of staging. When calculated with a power of 90%, the sample sizes increased by 35%. The number of years necessary to have enough patients for
The Clinical Respiratory Journal (2015) • ISSN 1752-6981 © 2015 John Wiley & Sons Ltd
Skaug et al.
valid estimates varies considerably between the four health regions given the number of incident cases in 2011. A management center with 200 patients needs 14 times as many years to sample enough patients for valid estimates of no staging than for valid estimates of 1-year survival.
Discussion For a group of chest physicians, the overall deviation from a national average of quality indicators required to cause a change in their management routines for lung cancer was around 20%. The deviation was slightly lower for 1-year survival than for unknown histology. Treatment differences of 10%–20% of outcome variables are often used in the calculation of sample sizes necessary to estimate effects in respiratory medicine clinical trials (19–22). In Norway, there are too many autonomous units managing lung cancer with too few patients. To generate valid and precise estimates for quality control, the units within a health region should work in a network with standardized programs for diagnosis and treatment. Accreditation as a lung cancer management center by Die Deutsche Krebsgesellschaft demands at least (i) 200 patients with lung cancer per year and (ii) 75 cases with surgical treatment (11). Our power calculations show that approximately 200 incident patients per year for 4 years will achieve valid estimates for the following quality indicators: staging, surgical resections and 1-year survival. This assumes that at least 90% of all incident patients are included in the registry. No statistical power calculations of the number of patients necessary to give precise estimates of quality of procedures of diagnosis or surgical treatment were available previously. The Norwegian government cancer strategy 2013–2017 recommends larger management units of care because this will improve survival and reduce complications and the recurrence of cancer (23). Previous studies of care have shown that a large volume of patients in a management unit is correlated with longer survival for breast cancer (24) and rectal cancer (25), but not significantly with resection of primary lung cancer (14). Approximately 200 incident patients per year for accreditation of a lung cancer center would allow establishment of one, one, two and six lung cancer management units in the health regions of North, Middle, West and South-East of Norway, respectively. Altogether, 10 or fewer lung cancer management centers would thus be available in Norway for valid comparisons. Indicators with a high percentage of
The Clinical Respiratory Journal (2015) • ISSN 1752-6981 © 2015 John Wiley & Sons Ltd
Quality, lung cancer, indicators, patient volume
events, such as 1-year survival, surgical resection and histology findings, could be used by all four health regions after inclusion of patients for 3 years. Only the South-East health region has the number of lung cancer patients necessary for comparison with national averages of the ‘no staging’ indicator. However, this indicator seems less useful as a quality indicator because as many as 95% of patients are staged. In the CRN, staging is defined as local, regional and distant (metastasis), and does not use the tumor–node– metastasis (TNM) staging system (I, II, III and IV). This renders impossible international comparisons (26). There is also little standardization and systematic quality control between management units and hospitals within a health region. Furthermore, in hospitals with many lung cancer patients, the diagnostic and therapeutic tasks are often not concentrated on a few persons to provide them with a supreme clinical experience. The national lung cancer quality registry should include the current quality indicators, as well as the updated TNM classification, performance status and documentation of all interventions (surgery, radiation, chemotherapy and molecular genetic treatment). All patients should be discussed with a multidisciplinary team at an accredited management center (27). A network like this for lung cancer care was developed successfully in Scotland (28). Palliative treatment could also be administered at smaller general hospitals, nursing homes and at home, providing standardized high-quality care supervised by a lung cancer management center, thus enabling the patients to have symptom control and end-of-life care in their own neighborhood and close to their family. In a prospective study, most terminal cancer patients wanted to die at home (29). Some of these changes in lung cancer care have started in Norway today. In larger hospitals, patients are given multidisciplinary assessments based on the diagnostic work done at smaller hospitals. This can be communicated via telemedicine, and thus the patients do not need to travel long distances until they eventually are referred to the larger hospitals for surgery and/or radiation treatment. Only research-based standardized procedures and treatments should be used within each of the accredited management units, with efficient communication with patients at home, in nursing institutions and in hospital. These changes in lung cancer care would enable the CRN to be an effective tool for surveillance of the quality of lung cancer care in Norway. A weakness is that the study among the chest physicians is performed in only one hospital. This is one of the largest chest hospitals in Norway, but the results 5
Quality, lung cancer, indicators, patient volume
could however have been different with a broader representation among the hospitals in Norway.
Conclusions Based on our calculations, we recommend that lung cancer management units within the Norwegian health regions should merge to include at least 200 incident lung cancer patients per year. This could give appropriate and precise estimates of quality indices of lung cancer care for comparisons between management centers.
Acknowledgments The study has used data from the Cancer Registry of Norway. The interpretation and reporting of these data are the sole responsibility of the authors, and no endorsement by the Cancer Registry of Norway is intended nor should be inferred. We thank Frøydis Langmark, MD, PhD, for comments on the early drafts of the manuscript. Furthermore, we want to thank the 26 physicians at the Department of Thoracic Medicine, Haukeland University Hospital for participating in the MID study.
References 1. Jakobsen E, Palshof T, Osterlind K, Pilegaard H. Data from a national lung cancer registry contributes to improve outcome and quality of surgery: Danish results. Eur J Cardiothorac Surg. 2009;35(2): 348–52. 2. Norsk_Lungekreft_gruppe. Retningslinjer for utredning og behandling av lungekreft. 2012. Available at: http://www.nlcgno/node/41 (accessed 09 August 2013). 3. Larsen IK, Smastuen M, Johannesen TB, et al. Data quality at the Cancer Registry of Norway: an overview of comparability, completeness, validity and timeliness. Eur J Cancer. 2009;45(7): 1218–31. 4. Barlow L, Westergren K, Holmberg L, Talback M. The completeness of the Swedish Cancer Register: a sample survey for year 1998. Acta Oncol. 2009;48(1): 27–33. 5. Rich AL, Tata LJ, Stanley RA, et al. Lung cancer in England: information from the National Lung Cancer Audit (LUCADA). Lung Cancer. 2011;72(1): 16–22. 6. Goeckenjan G, Sitter H, Thomas M, et al. [Prevention, diagnosis, therapy, and follow-up of lung cancer]. Pneumologie. 2010;64(Suppl. 2): e1–164. 7. Mainz J, Hansen AM, Palshof T, Bartels PD. National quality measurement using clinical indicators: the Danish National Indicator Project. J Surg Oncol. 2009;99(8): 500–4.
6
Skaug et al.
8. Sanchez de Cos J, Miravet L, Abal J, et al. Lung cancer survival in Spain and prognostic factors: a prospective, multiregional study. Lung Cancer. 2007;59(2): 246–54. 9. Myrdal G, Lamberg K, Lambe M, Stahle E, Wagenius G, Holmberg L. Regional differences in treatment and outcome in non-small cell lung cancer: a populationbased study (Sweden). Lung Cancer. 2009;63(1): 16–22. 10. von Plessen C, Strand TE, Wentzel-Larsen T, et al. Effectiveness of third-generation chemotherapy on the survival of patients with advanced non-small cell lung cancer in Norway: a national study. Thorax. 2008;63(10): 866–71. 11. Blum T, Schonfeld N, Kollmeier J, et al. [Lung cancer in Germany – the current state of management]. Pneumologie. 2011;65(1): 7–18. 12. Farjah F, Flum DR, Ramsey SD, Heagerty PJ, Symons RG, Wood DE. Multi-modality mediastinal staging for lung cancer among Medicare beneficiaries. J Thorac Oncol. 2009;4(3): 355–63. 13. Coory M, Fong K, Bowman R, Hall L. Why we need a population-based approach to clinical indicators for cancer: a case study using microscopic confirmation of lung cancer in Queensland. Intern Med J. 2006;36(6): 389–92. 14. Strand TE, Rostad H, Moller B, Norstein J. Survival after resection for primary lung cancer: a population based study of 3211 resected patients. Thorax. 2006;61(8): 710–5. 15. Larsen IK. Cancer in Norway 2011 – Cancer Incidence, Mortality, Survival and Prevalence in Norway. Oslo, Cancer Registry of Norway, 2013. 16. Helsedirektoratet. Nasjonale kvalitetsindikatorer Rapport N-020: Tid fra henvisning til første behandling for lungekreft. 2012. Available at: http://nesstar2.shdir.no/ kvalind/ (accessed 06 February 2014). 17. Skaug K, Eide GE, A. G. Lung cancer on Haugalandet, Norway. A long term follow up study. [Dissertation for the degree of philosopsiae doctor (PhD)], University of Bergen, 2012. 18. Jaeschke R, Singer J, Guyatt GH. Measurement of health status. Ascertaining the minimal clinically important difference. Control Clin Trials. 1989;10(4): 407–15. 19. Redelmeier DA, Bayoumi AM, Goldstein RS, Guyatt GH. Interpreting small differences in functional status: the six minute walk test in chronic lung disease patients. Am J Respir Crit Care Med. 1997;155(4): 1278–82. 20. Gulsvik A, Bakke P, Humerfelt S, Omenaas E, Baste V. Measurement of respiratory symptoms and sample size to detect a given difference between treatment groups in obstructive lung disease. Eur Repir Rew. 1991;1(5): 436–43. 21. Helbekkmo N, Sundstrom SH, Aasebo U, et al. Vinorelbine/carboplatin vs gemcitabine/carboplatin in advanced NSCLC shows similar efficacy, but different impact of toxicity. Br J Cancer. 2007;97(3): 283–9.
The Clinical Respiratory Journal (2015) • ISSN 1752-6981 © 2015 John Wiley & Sons Ltd
Skaug et al.
22. Hermes A, Bergman B, Bremnes R, et al. Irinotecan plus carboplatin versus oral etoposide plus carboplatin in extensive small-cell lung cancer: a randomized phase III trial. J Clin Oncol. 2008;26(26): 4261–7. 23. Ministry of Health and Care Services, editor. Sammen – Mot Kreft. Nasjonal Kreftstategi 2013–2017. Oslo, Norwegian Ministry of Health and Care Services, 2013. Available at: https://www.regjeringen.no/contentassets/ 07cd14ff763444a3997de1570b85fad1/kreftstrategien _2013.pdf (accessed 16 February 2015). 24. Scharl A, Gohring UJ. Does center volume correlate with survival from breast cancer? Breast Care. 2009;4(4): 237–44. 25. Wibe A, Eriksen MT, Syse A, et al. Effect of hospital caseload on long-term outcome after standardization
The Clinical Respiratory Journal (2015) • ISSN 1752-6981 © 2015 John Wiley & Sons Ltd
Quality, lung cancer, indicators, patient volume
26.
27. 28.
29.
of rectal cancer surgery at a national level. Br J Surg. 2005;92(2): 217–24. Walters S, Maringe C, Coleman MP, et al. Lung cancer survival and stage at diagnosis in Australia, Canada, Denmark, Norway, Sweden and the UK: a population-based study, 2004–2007. Thorax. 1990;68(6): 551–64. Detterbeck F. What is quality and does it matter? J Thorac Oncol. 2009;4(3): 279–80. Erridge SC, Murray B, Price A, et al. Improved treatment and survival for lung cancer patients in South-East Scotland. J Thorac Oncol. 2008;3(5): 491–8. Townsend J, Frank AO, Fermont D, et al. Terminal cancer care and patients’ preference for place of death: a prospective study. BMJ. 1990;301(6749): 415–7.
7
The Clinical Respiratory Journal
What is the minimum number of patients for quality control of lung cancer management in Norway? Knut Skaug1,2, Geir E. Eide3,4 and Amund Gulsvik2 1 Department of Medicine, NO-5504 Haugesund Hospital, Health Region of Fonna, Haugesund, Norway 2 Section of Thoracic Medicine, Department of Clinical Science, University of Bergen, Bergen, Norway 3 Centre for Clinical Research, Haukeland University Hospital, Bergen, Norway 4 Research Group for Lifestyle Epidemiology, Department of Global Public Health and Primary Care, University of Bergen, Bergen, Norway
Abstract Introduction: There are few data available on the optimal number of lung cancer patients needed to generate and compare estimates of quality between units managing lung cancer. The number of lung cancer patients per management unit varies considerably in Norway, where there are 42 hospitals that treated between 1 and 454 lung cancer patients in 2011. Aims: To estimate the differences in quality indicators that are of sufficient importance to change a pulmonary physician’s lung cancer management program, and to estimate the size of the patient samples necessary to detect such differences. Method: Twenty-six physicians were asked about the relative differences from a national average of quality indicators that would change their own lung cancer management program. Sample sizes were calculated to give valid estimates of quality of a management unit based on prevalence of quality indicators and minimally important differences (MID). Results: The average MID in quality indicators that would cause a change in management varied from 18% to 24% among 26 chest physicians, depending on the indicator. Conclusions: To generate precise estimates for quality control of lung cancer care in Norway, the number of management units must be reduced. Given the present willingness of chest physicians to change their procedures for management of lung cancer according to the results of quality control indicators, we recommend a maximum of 10 units with a minimum of 200 incident lung cancer patients per year for each management center. Please cite this paper as: Skaug K, Eide GE and Gulsvik A. What is the minimum number of patients for quality control of lung cancer management in Norway? Clin Respir J 2015; ••: ••–••. DOI:10.1111/crj.12274.
Introduction The quality of care in lung cancer is of major concern for patients and health-care providers (1). The purpose of national guidelines (2) and a quality registry is to improve the care of these patients (1). Few data are available on how national or regional management programs for lung cancer care are followed. The national cancer registries of the Nordic countries are
The Clinical Respiratory Journal (2015) • ISSN 1752-6981 © 2015 John Wiley & Sons Ltd
Key words lung cancer – indicators – patient volume – quality Correspondence Knut Skaug, MD, PhD, Department of Medicine, NO-5504 Haugesund Hospital, Health Region of Fonna, P.O. Box 2170, NO-5104 Haugesund, Norway. Tel: +47 5 273 20 00 Fax: +47 5 273 20 02 email: [email protected] Received: 27 March 2014 Revision requested: 17 October 2014 Accepted: 20 January 2015 DOI:10.1111/crj.12274 Authorship and contributorship Knut Skaug: planning of the project; collecting and analyzing the data; writing the manuscript. Geir E. Eide: analyzing the data; contributed to designing the study and manuscript. Amund Gulsvik: planning and designing of the project; writing the manuscript; senior author and adviser. Conflicts of interest The authors have stated explicitly that there are no conflicts of interest in connection with this article.
known to be of good quality (3, 4). Indicators used to compare, coordinate and improve lung cancer care have been established in England (5), Germany (6) and Denmark (7). Regional differences in survival within a country have been documented in Spain (8), Sweden (9) and Norway (10). However, little is known about how many patients a lung cancer management unit should treat per year to give meaningful estimates of quality (11). Quality
1
Quality, lung cancer, indicators, patient volume
Skaug et al.
indicators of lung cancer management that have been used are the proportion of patients staged through multimodal investigation (12), diagnosis verified by histology or cytology (13), proportion of patients offered surgery (14), and the percentage of patients with 1-year survival (15). The aims of our study were to estimate the differences in quality indicators that are of sufficient importance to change a pulmonary physician’s lung cancer management program, and to estimate the size of the patient samples necessary to detect such differences.
Materials and methods Organization of lung cancer care in Norway Since 2002, Norway has been divided into four health regions, North, Middle, West and South-East, which are given their yearly performance targets by the governmental Department of Health. The annual numbers of incident cases of lung cancer in 2011 in the four health regions were 291, 370, 508 and 1551, respectively (15). However, the autonomy of management within the health regions is considerable for hospitals, thoracic surgeons, oncologists and chest physicians. In 2011, treatment for lung cancer was provided by 42 of the 59 hospitals in Norway (Fig. 1). Altogether, 33 hospitals reported the time from acceptance of a referred patient to the first tumor-directed treatment, and the number of patients per management unit varied from 1 to 454 patients of a total of 1618 patients. Fourteen hospitals treated fewer than 10 patients with lung cancer per year (16). Resection surgery and/or radiation treatment were performed in 10 hospitals.
National cancer registry The population-based Cancer Registry of Norway (CRN) receives clinical and pathology reports for all
cancer patients in the country. Death certificates are an additional source of data. The CRN has been shown to include more than 98% of all cancer patients in the country (3).
Lung cancer quality indicators We used the same measures of diagnostic and treatment quality indicators as were used for the survey of health care for lung cancer in Haugalandet, Norway (17). To obtain the national values for these quality indicators, we got from CRN the number of all incident lung cancer patients in Norway, as described in Table 1. Then we got in percent (i) the proportion of the patients who were staged (2), (ii) the proportion with a histological verification of the diagnosis (13), (iii) the proportion of patients with surgical resections (14) and (iv) 1-year survival (15).
Minimally important differences (MID) in quality indicators The method for estimating MID has been described in several previous studies (18, 19). Twenty-six physicians at the Department of Thoracic Medicine, Haukeland University Hospital, Bergen, completed a questionnaire, giving the percentage difference from the national average for each of the quality indicators that would lead to changes in their own management of lung cancer. They were informed of the national average of the four quality indicators (Table 1). Each physician was asked to tick a box for predetermined relative differences (5%, 10%, 20%, 30%, 40%, 50%, 75% or 100%) from the national average that would lead them to change their own management program of procedures to diagnose and treat lung cancer. Fourteen of the chest physicians had
Table 1. Number of patients (sample size) required to detect a 20% relative deterioration from the observed national average for no histology*, no staging,† surgical resection† and 1-year survival† and years to collect sufficient patients in management units
Indicators of care
Observed national average (%)
20% relative deterioration
Sample size (number of patients)
No histology No staging Surgical resection One-year survival
23.9 5.0 17.9 38.9
28.7% 6.0% 14.3% 31.1%
531 3283 686 242
Health regions years to collect sample
Center
South-East (1551)‡
West (508)‡
Middle (370)‡
North (291)‡
(200)§
0.3 2.1 0.4 0.2
1.0 6.5 1.4 0.5
1.4 8.9 1.9 0.7
1.8 11.3 2.4 0.8
2.7 16.4 3.4 1.2
*Reported from Cancer Registry of Norway (CRN) on 2720 incident lung cancer patients in 2011 in Norway and the four health regions. †Reported from CRN on 7741 incident lung cancer patients 2008–2010. ‡Incident lung cancer patients in the four health regions in Norway in 2011. §Incident lung cancer patients to be given accreditation as lung cancer management center.
2
The Clinical Respiratory Journal (2015) • ISSN 1752-6981 © 2015 John Wiley & Sons Ltd
Skaug et al.
Quality, lung cancer, indicators, patient volume
Figure 1. Location of hospitals in Norway taking part in lung cancer care (2011).
The Clinical Respiratory Journal (2015) • ISSN 1752-6981 © 2015 John Wiley & Sons Ltd
3
Quality, lung cancer, indicators, patient volume
Skaug et al.
No histology (%)
10
10
20
20
30
30
40
40
50
50
No staging (%)
0
5
10
15
20
25
0
Mean 22.5 (line), 95% CI (16.3, 28.7) (dotted)
5
10
15
20
25
Mean 23.8 (line), 95% CI (16.8, 29.4) (dotted)
One-year survival (%)
0
5
10
15
20
25
Mean 18.9 (line), 95% CI (14.1, 23.6) (dotted)
40 30 20 10
10
20
30
40
One-year survival (%)
50
50
Surgical treatment (%)
0
5
10
15
20
25
Mean 17.9 (line), 95% CI (13.0, 22.8) (dotted)
Figure 2. Relative percentages deviation (minimally important difference) from the national average on four quality indicators sufficient for the physician to change their management of lung cancer. Answers were given by 26 physicians at the Department of Thoracic Medicine, University of Bergen. The x axis denotes physicians by number 1–26. CI, confidence interval.
4 years or more of training in a chest clinic, while 12 had less than 4 years of training. Using Cytel Studio (version 10.0; Cytel Corporation, Cambridge, MA, USA) and one-sided tests, we calculated the sample sizes required to obtain a preset deviation of 20% from the national average for the four quality indicators, with significance set at P < 0.05 and power at 0.80 and 0.90.
Results Deviation from the national average causing change of management The deviation in the proportion of patients with unknown stage that would induce change by the physicians in their lung cancer management was on average 23% and for unknown histology 24% (see Fig. 2). The relevant deviations in the percentage of
4
patients undergoing surgery and in 1-year survival were on average 19% and 18%, respectively. In a subanalysis comparing these results between physicians with 3 years of experience in thoracic medicine or less with those who had 4 years or more, there were no significant differences.
Necessary number of samples to detect a 20% deviation from national average The sample sizes required to detect a deviation of 20% for the four quality indicators compared with the national average with a power of 80% at a significance level of P ≤ 0.05 are shown in Table 1. The sample sizes varied from 242 patients for 1-year survival to 3283 patients for lack of staging. When calculated with a power of 90%, the sample sizes increased by 35%. The number of years necessary to have enough patients for
The Clinical Respiratory Journal (2015) • ISSN 1752-6981 © 2015 John Wiley & Sons Ltd
Skaug et al.
valid estimates varies considerably between the four health regions given the number of incident cases in 2011. A management center with 200 patients needs 14 times as many years to sample enough patients for valid estimates of no staging than for valid estimates of 1-year survival.
Discussion For a group of chest physicians, the overall deviation from a national average of quality indicators required to cause a change in their management routines for lung cancer was around 20%. The deviation was slightly lower for 1-year survival than for unknown histology. Treatment differences of 10%–20% of outcome variables are often used in the calculation of sample sizes necessary to estimate effects in respiratory medicine clinical trials (19–22). In Norway, there are too many autonomous units managing lung cancer with too few patients. To generate valid and precise estimates for quality control, the units within a health region should work in a network with standardized programs for diagnosis and treatment. Accreditation as a lung cancer management center by Die Deutsche Krebsgesellschaft demands at least (i) 200 patients with lung cancer per year and (ii) 75 cases with surgical treatment (11). Our power calculations show that approximately 200 incident patients per year for 4 years will achieve valid estimates for the following quality indicators: staging, surgical resections and 1-year survival. This assumes that at least 90% of all incident patients are included in the registry. No statistical power calculations of the number of patients necessary to give precise estimates of quality of procedures of diagnosis or surgical treatment were available previously. The Norwegian government cancer strategy 2013–2017 recommends larger management units of care because this will improve survival and reduce complications and the recurrence of cancer (23). Previous studies of care have shown that a large volume of patients in a management unit is correlated with longer survival for breast cancer (24) and rectal cancer (25), but not significantly with resection of primary lung cancer (14). Approximately 200 incident patients per year for accreditation of a lung cancer center would allow establishment of one, one, two and six lung cancer management units in the health regions of North, Middle, West and South-East of Norway, respectively. Altogether, 10 or fewer lung cancer management centers would thus be available in Norway for valid comparisons. Indicators with a high percentage of
The Clinical Respiratory Journal (2015) • ISSN 1752-6981 © 2015 John Wiley & Sons Ltd
Quality, lung cancer, indicators, patient volume
events, such as 1-year survival, surgical resection and histology findings, could be used by all four health regions after inclusion of patients for 3 years. Only the South-East health region has the number of lung cancer patients necessary for comparison with national averages of the ‘no staging’ indicator. However, this indicator seems less useful as a quality indicator because as many as 95% of patients are staged. In the CRN, staging is defined as local, regional and distant (metastasis), and does not use the tumor–node– metastasis (TNM) staging system (I, II, III and IV). This renders impossible international comparisons (26). There is also little standardization and systematic quality control between management units and hospitals within a health region. Furthermore, in hospitals with many lung cancer patients, the diagnostic and therapeutic tasks are often not concentrated on a few persons to provide them with a supreme clinical experience. The national lung cancer quality registry should include the current quality indicators, as well as the updated TNM classification, performance status and documentation of all interventions (surgery, radiation, chemotherapy and molecular genetic treatment). All patients should be discussed with a multidisciplinary team at an accredited management center (27). A network like this for lung cancer care was developed successfully in Scotland (28). Palliative treatment could also be administered at smaller general hospitals, nursing homes and at home, providing standardized high-quality care supervised by a lung cancer management center, thus enabling the patients to have symptom control and end-of-life care in their own neighborhood and close to their family. In a prospective study, most terminal cancer patients wanted to die at home (29). Some of these changes in lung cancer care have started in Norway today. In larger hospitals, patients are given multidisciplinary assessments based on the diagnostic work done at smaller hospitals. This can be communicated via telemedicine, and thus the patients do not need to travel long distances until they eventually are referred to the larger hospitals for surgery and/or radiation treatment. Only research-based standardized procedures and treatments should be used within each of the accredited management units, with efficient communication with patients at home, in nursing institutions and in hospital. These changes in lung cancer care would enable the CRN to be an effective tool for surveillance of the quality of lung cancer care in Norway. A weakness is that the study among the chest physicians is performed in only one hospital. This is one of the largest chest hospitals in Norway, but the results 5
Quality, lung cancer, indicators, patient volume
could however have been different with a broader representation among the hospitals in Norway.
Conclusions Based on our calculations, we recommend that lung cancer management units within the Norwegian health regions should merge to include at least 200 incident lung cancer patients per year. This could give appropriate and precise estimates of quality indices of lung cancer care for comparisons between management centers.
Acknowledgments The study has used data from the Cancer Registry of Norway. The interpretation and reporting of these data are the sole responsibility of the authors, and no endorsement by the Cancer Registry of Norway is intended nor should be inferred. We thank Frøydis Langmark, MD, PhD, for comments on the early drafts of the manuscript. Furthermore, we want to thank the 26 physicians at the Department of Thoracic Medicine, Haukeland University Hospital for participating in the MID study.
References 1. Jakobsen E, Palshof T, Osterlind K, Pilegaard H. Data from a national lung cancer registry contributes to improve outcome and quality of surgery: Danish results. Eur J Cardiothorac Surg. 2009;35(2): 348–52. 2. Norsk_Lungekreft_gruppe. Retningslinjer for utredning og behandling av lungekreft. 2012. Available at: http://www.nlcgno/node/41 (accessed 09 August 2013). 3. Larsen IK, Smastuen M, Johannesen TB, et al. Data quality at the Cancer Registry of Norway: an overview of comparability, completeness, validity and timeliness. Eur J Cancer. 2009;45(7): 1218–31. 4. Barlow L, Westergren K, Holmberg L, Talback M. The completeness of the Swedish Cancer Register: a sample survey for year 1998. Acta Oncol. 2009;48(1): 27–33. 5. Rich AL, Tata LJ, Stanley RA, et al. Lung cancer in England: information from the National Lung Cancer Audit (LUCADA). Lung Cancer. 2011;72(1): 16–22. 6. Goeckenjan G, Sitter H, Thomas M, et al. [Prevention, diagnosis, therapy, and follow-up of lung cancer]. Pneumologie. 2010;64(Suppl. 2): e1–164. 7. Mainz J, Hansen AM, Palshof T, Bartels PD. National quality measurement using clinical indicators: the Danish National Indicator Project. J Surg Oncol. 2009;99(8): 500–4.
6
Skaug et al.
8. Sanchez de Cos J, Miravet L, Abal J, et al. Lung cancer survival in Spain and prognostic factors: a prospective, multiregional study. Lung Cancer. 2007;59(2): 246–54. 9. Myrdal G, Lamberg K, Lambe M, Stahle E, Wagenius G, Holmberg L. Regional differences in treatment and outcome in non-small cell lung cancer: a populationbased study (Sweden). Lung Cancer. 2009;63(1): 16–22. 10. von Plessen C, Strand TE, Wentzel-Larsen T, et al. Effectiveness of third-generation chemotherapy on the survival of patients with advanced non-small cell lung cancer in Norway: a national study. Thorax. 2008;63(10): 866–71. 11. Blum T, Schonfeld N, Kollmeier J, et al. [Lung cancer in Germany – the current state of management]. Pneumologie. 2011;65(1): 7–18. 12. Farjah F, Flum DR, Ramsey SD, Heagerty PJ, Symons RG, Wood DE. Multi-modality mediastinal staging for lung cancer among Medicare beneficiaries. J Thorac Oncol. 2009;4(3): 355–63. 13. Coory M, Fong K, Bowman R, Hall L. Why we need a population-based approach to clinical indicators for cancer: a case study using microscopic confirmation of lung cancer in Queensland. Intern Med J. 2006;36(6): 389–92. 14. Strand TE, Rostad H, Moller B, Norstein J. Survival after resection for primary lung cancer: a population based study of 3211 resected patients. Thorax. 2006;61(8): 710–5. 15. Larsen IK. Cancer in Norway 2011 – Cancer Incidence, Mortality, Survival and Prevalence in Norway. Oslo, Cancer Registry of Norway, 2013. 16. Helsedirektoratet. Nasjonale kvalitetsindikatorer Rapport N-020: Tid fra henvisning til første behandling for lungekreft. 2012. Available at: http://nesstar2.shdir.no/ kvalind/ (accessed 06 February 2014). 17. Skaug K, Eide GE, A. G. Lung cancer on Haugalandet, Norway. A long term follow up study. [Dissertation for the degree of philosopsiae doctor (PhD)], University of Bergen, 2012. 18. Jaeschke R, Singer J, Guyatt GH. Measurement of health status. Ascertaining the minimal clinically important difference. Control Clin Trials. 1989;10(4): 407–15. 19. Redelmeier DA, Bayoumi AM, Goldstein RS, Guyatt GH. Interpreting small differences in functional status: the six minute walk test in chronic lung disease patients. Am J Respir Crit Care Med. 1997;155(4): 1278–82. 20. Gulsvik A, Bakke P, Humerfelt S, Omenaas E, Baste V. Measurement of respiratory symptoms and sample size to detect a given difference between treatment groups in obstructive lung disease. Eur Repir Rew. 1991;1(5): 436–43. 21. Helbekkmo N, Sundstrom SH, Aasebo U, et al. Vinorelbine/carboplatin vs gemcitabine/carboplatin in advanced NSCLC shows similar efficacy, but different impact of toxicity. Br J Cancer. 2007;97(3): 283–9.
The Clinical Respiratory Journal (2015) • ISSN 1752-6981 © 2015 John Wiley & Sons Ltd
Skaug et al.
22. Hermes A, Bergman B, Bremnes R, et al. Irinotecan plus carboplatin versus oral etoposide plus carboplatin in extensive small-cell lung cancer: a randomized phase III trial. J Clin Oncol. 2008;26(26): 4261–7. 23. Ministry of Health and Care Services, editor. Sammen – Mot Kreft. Nasjonal Kreftstategi 2013–2017. Oslo, Norwegian Ministry of Health and Care Services, 2013. Available at: https://www.regjeringen.no/contentassets/ 07cd14ff763444a3997de1570b85fad1/kreftstrategien _2013.pdf (accessed 16 February 2015). 24. Scharl A, Gohring UJ. Does center volume correlate with survival from breast cancer? Breast Care. 2009;4(4): 237–44. 25. Wibe A, Eriksen MT, Syse A, et al. Effect of hospital caseload on long-term outcome after standardization
The Clinical Respiratory Journal (2015) • ISSN 1752-6981 © 2015 John Wiley & Sons Ltd
Quality, lung cancer, indicators, patient volume
26.
27. 28.
29.
of rectal cancer surgery at a national level. Br J Surg. 2005;92(2): 217–24. Walters S, Maringe C, Coleman MP, et al. Lung cancer survival and stage at diagnosis in Australia, Canada, Denmark, Norway, Sweden and the UK: a population-based study, 2004–2007. Thorax. 1990;68(6): 551–64. Detterbeck F. What is quality and does it matter? J Thorac Oncol. 2009;4(3): 279–80. Erridge SC, Murray B, Price A, et al. Improved treatment and survival for lung cancer patients in South-East Scotland. J Thorac Oncol. 2008;3(5): 491–8. Townsend J, Frank AO, Fermont D, et al. Terminal cancer care and patients’ preference for place of death: a prospective study. BMJ. 1990;301(6749): 415–7.
7
