Hybrid analysis for up-to-date long-term survival rates in cancer registries with delayed recording of incident cases

Up-to-date monitoring of long-term survival is an important task of population-based and clinical cancer registries. A few years ago, a new method of survival analysis, denoted period analysis, was introduced to provide more up-to-date estimates of long-term survival. However, a prototypical period analysis may not be applicable in situations with delayed recording of incident cases. We introduce herein a hybrid type of analysis that combines elements of both traditional and period analyses which may still be feasible in such settings. The performance of the hybrid type of analysis compared with other design options is empirically evaluated and illustrated for children diagnosed with cancer in the United States. The empirical evaluation indicates that hybrid analysis may be useful to derive more up-to-date estimates of long-term survival compared with traditional design options if there is a strong improvement of survival over time, even in situations with a substantial delay in recording of incident cases.     

Maximizing the benefits of model-based period analysis of cancer patient survival.

Period analysis has been shown to provide more up-to-date estimates of cancer survival than traditional methods of survival analysis. There is, however, a tradeoff between up-to-dateness and precision of period survival estimates: increasing up-to-dateness by restricting the analysis to a relatively short period, such as the most recent calendar year, goes along with loss of precision. Recently, a model-based approach was proposed, in which more precise period survival estimates for the most recent year can be obtained through modeling of survival trends within a recent 5-year period. We assess possibilities to extend the time window used for modeling to come up with even more precise, but equally accurate and up-to-date estimates of prognosis. Empirical evaluation using data from the Finnish Cancer Registry shows that extension of the time window to about 10 years provides, in most cases, as accurate results as using a 5-year time window (whereas further extension may lead to considerably less accurate results in some cases). Using 10-year time windows for modeling, SEs of survival estimates can be approximately halved compared with conventional period survival estimates for the most recent calendar year. Furthermore, we present a modification of the modeling approach, which allows extension to 10-year time windows to be achieved without the need to include additional cohorts of patients diagnosed longer time ago and which provides similarly accurate survival estimates at comparable levels of precision in most cases. Our analyses indicate opportunities to further maximize benefits of model-based period analysis of cancer survival.     

Further enhanced monitoring of cancer patient survival by stage-adjusted period analysis.

Monitoring progress in cancer patient survival is an important task of population-based cancer registration. Period analysis has been shown to provide more up-to-date estimates of cancer patient survival than traditional methods of survival analysis. However, even period estimates may disclose recent improvements in long-term survival with some delay as they are still partly based on the survival experience of patients diagnosed years ago. If these patients had a less favorable stage distribution than the patients diagnosed in a more recent calendar period (e.g., due to progress in early detection), period estimates may underestimate long-term survival for patients diagnosed in that period. This particular source of potential underestimation can be overcome by adjustment of the stage distribution of all patients included in period analysis to the stage distribution of the patients diagnosed in the period of interest. The principle, application, and use of stage adjustment of period survival estimates are illustrated with 5- and 10-year relative survival estimates of patients diagnosed with breast cancer and followed with respect to survival in the United States between 1973 and 2001, using data of the Surveillance, Epidemiology, and End Results Program of the National Cancer Institute. We show that stage adjustment may often further enhance the benefits of period analysis for deriving up-to-date cancer survival estimates.     

Long-term cancer patient survival achieved by the end of the 20th century: most up-to-date estimates from the nationwide Finnish cancer registry

A new method of survival analysis, denoted period analysis, has recently been developed, which has been shown to provide more up-to-date estimates of long-term survival rates than traditional methods of survival analysis. We applied period analysis to data from the nationwide Finnish cancer registry to provide up-to-date estimates of 5-, 10-, 15- and 20-year relative survival rates (RSR) achieved by the end of the 20th century. For most forms of cancer, period estimates of long-term survival are much higher than corresponding traditional survival estimates which suggests that for these cancers there has been ongoing major progress in survival rates in recent years which so far has remained undisclosed by traditional methods of survival analysis. For example, period analysis reveals that 10 year RSR have come close to (or even exceed) 80% for cancer of the corpus uteri and melanoma, 75% for breast cancer, 70% for bladder cancer, 65% for cancer of the cervix uteri, and 55% for cancer of the colon and prostate. Period analysis further reveals that 20 year RSR have now come close to (or even exceed) 75% for endometrial cancer and melanoma, 60% for breast cancer and cervical cancer, 55% for colon cancer and bladder cancer, and 40%-50% for cancer of the rectum, the ovaries, kidneys and nervous system.     

A computer program for period analysis of cancer patient survival

Monitoring of long-term survival rates, which is now routinely performed by many cancer registries throughout the world, should be as up-to-date as possible. A few years ago, a new method of survival analysis, denoted period analysis, has been proposed which provides more up-to-date estimates of long-term survival rates than traditional survival analysis by exclusively reflecting the survival experience of patients within a recent calendar period. However, application of this method has so far been hindered by the lack of pertinent computer programs. In this paper, we present a simple and easy-to-use computer program (SAS macro) that enables one to carry out period analysis (as well as conventional analysis) of both absolute and relative survival rates with the type of data commonly available in population-based cancer registries. We illustrate application of the program with examples from the nationwide Finnish Cancer Registry.     

Up-to-date long-term survival curves of patients with cancer by period analysis.

PURPOSE: Provision of up-to-date long-term survival curves is an important task of cancer registries. Traditionally, survival curves have been derived for cohorts of patients diagnosed many years ago. Using data of the Finnish Cancer Registry, we provide an empirical assessment of the use of a new method of survival analysis, denoted period analysis, for deriving more up-to-date survival curves. PATIENTS AND METHODS: We calculated 10-year relative survival curves actually observed for patients diagnosed with one of the 15 most common forms of cancer in 1983 to 1987, and we compared them with the most up-to-date 10-year relative survival curves that might have been obtained in 1983 to 1987 using either traditional (cohort-wise) or period analysis. We also give the most recent 10-year survival curves obtained by period analysis for the 1993 to 1997 period. RESULTS: For all forms of cancer, period analysis of the 1983 to 1987 data yielded survival curves that were very close to the survival curves later observed for patients who were newly diagnosed in that period (median and maximum difference of 10-year relative survival estimates: 0.9 and 5.7 percent units, respectively). By contrast, the survival curves obtained by traditional (cohort-wise) survival analysis in 1983 to 1987 would have been much lower for most forms of cancer (median and maximum difference: 5.8 and 18.4 percent units, respectively). The 10-year survival curves for the 1993 to 1997 period are substantially more favorable than previously available, traditionally derived survival curves for most forms of cancer. CONCLUSION: Period analysis is a useful tool for deriving up-to-date long-term survival curves of patients with cancer.     

Period versus cohort modeling of up-to-date cancer survival

Recently, 2 modeling strategies have been proposed and shown to be useful to increase precision of up-to-date cancer survival estimates and to predict cancer patient survival: modeled period analysis and modeled cohort analysis. We aimed to compare the performance of both types of modeling for providing up-to-date and precise cancer survival estimates. Data from the nationwide Finnish Cancer Registry were used to assess how well both approaches would have been able to predict 5-year relative survival of concurrently diagnosed patients if they had been applied for that purpose throughout the past decades. Analyses were carried out for 20 common forms of cancer. For each cancer, 5-year relative survival was modeled with either approach for each single calendar year from 1962 to 1997. Mean differences and mean squared differences from 5-year relative survival later observed for patients diagnosed in the 5-year period around those calendar years were calculated. Survival estimates obtained by period modeling had much lower standard errors than those obtained by cohort modeling. Furthermore, for a clear majority of cancers, period modeling on average also provided better prediction of 5-year relative survival than cohort modeling. We conclude that, although both modeling strategies have their merits and specific indications, period modeling of survival has distinct advantages for up-to-date and precise estimation of cancer survival in population-based cancer survival studies.     

Up-to-date long-term survival of cancer patients: an evaluation of period analysis on Swedish Cancer Registry data

The natural development of cancers as well as the measures to fight the disease are often long processes that require decades of follow up. Available information on long-term survival will thus often appear outdated and irrelevant. A few years ago, period-survival analysis was proposed as a means to obtain more up-to-date information on long-term cancer survival. This article assesses period and conventional cohort-based survival analyses on their ability to predict future survival. Based on historical data from the nationwide Swedish Cancer Registry 5-, 10- and 15-year relative survival actually observed for patients diagnosed at one particular point in time are compared to the most recent period and cohort-based survival estimates available at that point in time. The study shows that period analysis can, in most cases, be used to provide more up-to-date long-term estimates of cancer survival. Period analysis reduces the time lag of the survival estimates by some 5-10 years for all cancers combined and especially affects the survival estimates for small intestine carcinoids, meningioma and intracranial neurinoma of the brain, non-seminoma testicular cancer, chronic lymphocytic leukaemia and Hodgkin's lymphoma.     

Up-to-date survival curves of children with cancer by period analysis

Survival rates of children with cancer have strongly improved during the past decades, but much of this improvement has been disclosed with substantial delay by traditional methods of survival analysis, which reflect survival experience of patients diagnosed many years ago. In this paper, the use of a new method of survival analysis, denoted period analysis, for providing more up-to-date estimates of 10-year survival curves of children with cancer is empirically evaluated using data of the Surveillance, Epidemiology, and End Results Program of the United States National Cancer Institute. It is shown that period analysis provides much more up-to-date estimates of survival curves than traditional cohort-based survival analysis indeed, at least as long as there is ongoing improvement in survival rates over time, as it seems to be the case for many forms of childhood cancer. The most recent 10-year period survival estimates indicate that survival rates of children with cancer achieved by the end of the 20th century are substantially higher than previously available survival statistics have suggested. Application of period analysis may be particularly useful in the field of childhood cancer as it may help to prevent patients, their families and clinicians from being burdened by outdated, often too pessimistic survival expectations.     

Timely disclosure of progress in childhood cancer survival by 'period' analysis in the Automated Childhood Cancer Information System.

BACKGROUND: A few years ago, a new method of survival analysis, denoted 'period' analysis, was introduced to provide more up-to-date survival estimates of cancer patients. PATIENTS AND METHODS: We evaluated the period survival method using the large database of the Automated Childhood Cancer Information System (ACCIS). Our evaluation is based on data from 35 191 children diagnosed with cancer in 13 European countries between 1975 and 1989 and followed for vital status until around 1999. RESULTS: Using the follow-up data available in 1989, 10-year survival for all children with cancer calculated by the period method for the 1985-89 period was 58%, while it was 43% when calculated by traditional 'cohort' life-table analysis (based on children diagnosed in 1975-79). The period method provided a better estimate of the true 10-year survival of 62%, observed 10 years later in the cohort of patients diagnosed in 1985-89. Similar results were observed for each of the common groups of childhood cancer. CONCLUSION: Period analysis is especially useful for monitoring childhood cancer survival, because at a given point in time it provides more timely estimates of long-term survival expectations than the cohort life-table method. Using the ACCIS database, up-to-date estimates of period survival for childhood cancer are derived in subsequent papers in this journal.     

Up‐to‐date cancer survival: Period analysis and beyond

Since its introduction in 1996, period analysis has been shown to be useful for deriving more up‐to‐date cancer survival estimates, and the method is now increasingly used for that purpose in national and international cancer survival studies. However, period analysis, like other commonly employed methods, is just a special case from a broad class of design options in the analysis of cancer survival data. Here, we explore a broader range of design options, including 2 model‐based approaches, for deriving up‐to‐date estimates of 5‐ and 10‐year relative survival for patients diagnosed in the most recent 5‐year interval for which data are available. The performance of the various designs is evaluated empirically for 20 common forms of cancer using more than 50‐year long time series of data from the Finnish Cancer Registry. Period analysis as well as the 2 model‐based approaches, one using a “cohort‐type model” and another using a “period‐type model”, all performed better than traditional cohort or complete analysis. Compared with “standard period analysis”, the cohort‐type model further increased up‐to‐dateness of survival estimates, whereas the period‐type model increased their precision. While our analysis confirms advantages of period analysis over traditional methods in terms of up‐to‐dateness of cancer survival data, further improvements are possible by flexible use of model‐based approaches. © 2008 Wiley‐Liss, Inc.     

Long-term survival expectations of cancer patients in Europe in 2000–2002

Period analysis has been shown to provide more up-to-date estimates of long-term cancer survival rates than traditional cohort-based analysis. Here, we provide detailed period estimates of 5- and 10-year relative survival by cancer site, country, sex and age for calendar years 2000–2002. In addition, pan-European estimates of 1-, 5- and 10-year relative survival are provided. Overall, survival estimates were mostly higher than previously available cohort estimates. For most cancer sites, survival in countries from Northern Europe, Central Europe and Southern Europe was substantially higher than in the United Kingdom and Ireland and in countries from Eastern Europe. Furthermore, relative survival was also better in female than in male patients and decreased with age for most cancer sites.     

Up-to-date estimates of cancer patient survival even with common latency in cancer registration.

In an era of ongoing improvement in cancer patient survival, available long-term survival figures from cancer registries are often outdated and too pessimistic for two reasons: first, delay in availability of cancer registry data, typically in the order of a few years, and, second, application of cohort-based methods of survival analysis, which provide survival estimates for patients diagnosed many years ago. We developed a model-based period analysis approach aimed to overcome both problems. We provide extensive empirical evaluation of our approach by comparing its performance with that of previously available methods for monitoring of 5- and 10-year relative survival, with the use of data from the nationwide Finnish Cancer Registry of 490,279 patients ages >/=15 years and diagnosed with one of 20 common forms of cancer between 1953 and 1997. We show that, in most cases, the model-based approach predicts 5- and 10-year relative survival expectations of newly diagnosed patients quite closely and much better than any of the previously available methods, including standard period analysis. We conclude that the model-based approach may enable deriving up-to-date cancer survival rates even with the common latency in availability of cancer registry data.     

Site-specific proportion cured models applied to cancer registry data

Purpose

Proportion-cured models were applied to evaluate their applicability on data from a relatively small cancer registry and to assess the up-to-date survival level of major cancer types in Tyrol, Austria.

Methods

In total, the 25 most common types of cancer were analyzed with mixture cure models using the period approach for estimation of the proportion cured and median survival time of the fatal cases.

Results

For several of the cancer types, no estimates could be obtained. The models converged for 14 sites among females and for 15 among males. The highest estimate of the proportion cured was found for cervix cancer (74.0 %; 95 % CI 64.4–83.6) and the lowest for male pancreas cancer (4.6 %; 95 % CI 0.2–9.0). The highest median survival of the uncured was 2.7 years (95 % CI 1.2–6.0) for male larynx cancer and the lowest 0.3 years (95 % CI 0.1–0.6) for male acute myeloblastic leukemia (AML).

Conclusions

The estimates seem reliable for stomach, colon, rectum, pancreas, lung, cervix, ovary, central nervous system/brain and AML cancer and among men also for head/neck, esophagus, liver and kidney cancer. Altogether, it is demonstrated that even data from a regional cancer registry covering a rather small region can be utilized to derive up-to-date survival estimates of various cancer types, enabling monitoring of the development and changes in cancer treatment. Moreover, potentially this methodology is advantageously employable in any situation where the number of cancer cases is limited.     

Survival from common and rare cancers in Germany in the early 21st century

BackgroundUntil recently, population-based data of cancer survival in Germany mostly relied on one registry covering ∼1 million people (1.3% of the German population). Here, we provide up-to-date cancer survival estimates for Germany based on data from 11 population-based cancer registries, covering 33 million people and compare them to survival estimates from the United States.Patients and methodsCancer patients diagnosed in 1997–2006 were included. Period analysis was employed to calculate 5-year relative survival for 38 cancers for 2002–2006. German and USA survival rates were compared utilizing the Surveillance, Epidemiology and End Results 13 database.ResultsFive-year relative survival >80% was observed for testicular cancer (93.5%), skin melanoma (89.4%), cancers of the prostate (89.1%) and thyroid (87.8%), Hodgkin’s lymphoma (84.5%) and cancers of the breast (83.7%) and endometrium (81.0%), which together account for almost 40% of cases. For the majority of cancers, German survival estimates were close to or below those in the United States. Exceptions with higher survival in Germany were cancers of the stomach, pancreas and kidney and Hodgkin’s lymphoma.ConclusionsGerman cancer survival estimates are mostly higher than the 2000–2002 pan-European estimates. Further research is needed to investigate causes responsible for differences between German and USA cancer survival rates.     

Expected long-term survival of patients diagnosed with acute myeloblastic leukemia during 2006–2010

BackgroundTreatment of acute myeloblastic leukemia (AML) has evolved over the past several decades. Therefore, currently available estimates of long-term survival, which are based on survival for patients treated with potentially now obsolete protocols, may not pertain to patients currently diagnosed.MethodsUsing data from the 1973–2005 database of the Surveillance, Epidemiology, and End Results Program, we empirically validated a novel model-based method to project 5- and 10-year relative survival of AML patients and we applied the method to project relative survival of AML patients in the United States diagnosed during 2006–2010.ResultsEmpirical evaluation indicated that the modeling approach provides more accurate estimates of currently diagnosed patients than standard methods of survival analysis, such as cohort analysis or period analysis, in the majority of cases. Projected figures for 2006–2010 show 5- and 10-year relative survival estimates of 21.4% and 18.7% for all ages combined, 62.2% and 57.4% for ages 25–34, and 60.6% and 58.1% for ages 35–44. These estimates are substantially higher than the most up-to-date estimates obtained by standard survival analysis.ConclusionPatients diagnosed with AML during 2006–2010 at younger ages have much higher long-term survival expectations than indicated by previously available survival statistics.     

Assessment of long-term survival of cancer patients using cancer registry data from eastern China: Period analysis is superior to traditional methods

We aimed to provide a systematical evaluation of the performance of period analysis compared to traditional cohort and complete methods, using cancer registry data from Taizhou, eastern China. Overall, 5-year relative survival (RS) estimate was calculated using cohort analysis, complete analysis and period analysis, respectively; further analyses were stratified by sex, region, age at diagnosis and cancer sites. Deviation value (DV), defined as the deviation between the estimated 5-year RS obtained from each method and the observed actual survival, was calculated to evaluate the accuracy of each method. Overall, 5-year RS derived by period analysis were much closer to the observed actual survival (51.4%), compared to those by complete and cohort methods, with the estimates of 48.7% (DV: −2.7%), 43.2% (DV: −8.2%) and 36.3% (DV: −15.1%), respectively. Further stratifications by sex, age at diagnosis, region and cancer sites also supported period analysis provided more precise estimates, compared to complete and cohort methods. We found, for first time systematically using cancer registry data from eastern China, period analysis provided more up-to-date precise estimates of long-term survival for overall and stratifications by sex, age at diagnosis, region and cancer sites, compared to traditional cohort and complete methods. Nevertheless, further investigations using large cancer registry data across China are warranted for the widespread use of period analysis in China.     

Interpreting international comparisons of cancer survival: the effects of incomplete registration and the presence of death certificate only cases on survival estimates

We have assessed the impact on survival estimates based on cancer registry data of incomplete ascertainment of cancer cases and the presence of cases registered purely from death certificate information (DCO cases). Using data from the Thames and Finnish Cancer Registries we obtained five-year relative survival estimates for 12 cancer sites, excluding DCOs as usual. We then made adjustments to allow for the effects of both the known proportion of DCOs and the estimated proportion of missing cases for each site. In general, adjusting for DCOs led to lower survival estimates, whilst adjusting for incompleteness had the opposite effect. The Finnish data were largely complete and had small proportions of DCOs, and hence the adjustments had little effect on estimated survival. The changes in the Thames estimates were more marked. When performing cohort survival analysis (based on diagnoses between 1990 and 1994), the increases in the survival estimates gained from adjusting for incompleteness were for the most part offset by the decrease produced when adjusting for DCOs. However, when performing period survival analysis based on the period 1997-2001 (when the DCO rate at Thames had fallen by around a half relative to the earlier period), the final estimates (adjusted for both effects) were generally higher than the unadjusted values--thus reducing the apparent difference between the two countries. It is important to take variations in DCO proportion and/or completeness into consideration when comparing survival estimates between different populations.     

Disclosing progress in cancer survival with less delay

Cancer registration plays a key role in monitoring the burden of cancer. However, cancer registry (CR) data are usually made available with substantial delay to ensure best possible completeness of case ascertainment. Here, we investigate empirically with routinely available data whether such a delay is mandatory for survival analyses or whether data can be used earlier to provide more up-to-date survival estimates. We compared distributions of prognostic factors and period relative survival estimates for three population-based CRs in Germany (Schleswig-Holstein (SH), Rhineland-Palatinate (RP), Saarland (SA)) computed on datasets extracted one (DY+1) to 5 years after the year of diagnosis (DY+5; reference). Analyses were conducted for seven cancer sites and various survival analyses scenarios. The proportion of patients registered in the datasets at a given time varied strongly across registries with 57% (SH), 2% (RP) and 26% (SA) registered in DY+1 and >93% in all registries in DY+3. Five-year survival estimates for the most recent three-year period were comparable to estimates from the reference dataset already in DY+1 (mean absolute deviations = 0.2–0.6% units). Deviations >1% units were only observed for pancreatic and lung cancer in RP and leukemia in SA (all ≤1.5% units). For estimates of 1-year survival based on the most recent 1-year period only, slightly longer delays were required, but reasonable estimates were still obtained after 1–2 years, depending on the CR and cancer site. Thus, progress in cancer survival could be disclosed in a more timely manner than commonly practiced despite delays in completeness of registration.     

Long-term survival of children with leukemia achieved by the end of the second millennium.

BACKGROUND: The prognosis for patients with childhood leukemia has improved steadily over the last decades due to major progress in therapy. Much of this progress remains unaccounted for in traditional estimates of long-term survival rates, which essentially reflect the survival experience of patients who were diagnosed many years ago. METHODS: The authors applied a new method of survival analysis, called period analysis, to provide up-to-date estimates of long-term survival rates. The analysis is based on data from the nationwide German Childhood Cancer Registry and includes 8059 children who were diagnosed with leukemia between 1981 and 1998. The most up-to-date 5-year, 10-year, and 15-year survival estimates were obtained by period analysis and were compared with to the most up-to-date survival estimates from traditional methods of survival analysis. RESULTS: Period estimates (95% confidence intervals) of 5-year, 10-year, and 15-year survival rates achieved by 1998 were 81% (79-83%), 77% (74-79%), and 73% (70-76%), respectively, for all patients with leukemia combined; 86% (84-88%), 81% (79-84%), and 77% (74-81%), respectively, for patients with acute lymphocytic leukemia; and 59% (53-65%), 59% (53-65%), and 57% (49-64%), respectively, for patients with acute nonlymphocytic leukemia. Substantially lower estimates would have been obtained with traditional methods of survival analysis. CONCLUSIONS: These results from one of the world's largest childhood cancer registries reveal that cure rates of childhood leukemia achieved by the end of the second millennium are higher than suggested by previous estimates based on traditional methods of survival analysis.