Betty J. Flehinger

The Memorial Sloan-Kettering lung cancer screening program was begun in 1974 to evaluate sputum cytology as a supplement to the annual chest x-ray examination for early detection and diagnosis. The 10,040 adult, male cigarette smokers who enrolled were randomly assigned to receive annual chest x-ray examinations only or a dual screen with annual chest x-ray examination and four monthly sputum cytology evaluation. Over 40 percent of the 288 who developed lung cancer were diagnosed in stage I, and their survival was 76 percent at five years; overall survival was 35 percent. Nearly one third of the lung cancers detected on first examination on the dual screen, and 14 percent of those on subsequent examinations were found by cytologic examination. The same number of cancers developed in the x-ray screen only group, and were diagnosed at a later date. Despite the delay, survival and mortality were the same, suggesting that the squamous carcinomas detected by cytologic examination alone are very slow growing and tend to remain localized until detectable by x-ray examination.

Whereas most physicians believe that long-term survival is unlikely when mediastinal lymph node metastases are present, a significant number of these patients do have resectable tumors with encouraging long-term survival results. Data are presented to support this view, and steps identified to guide the physicians in selecting the patients who can benefit from this surgical approach.

The high incidence of and mortality from colorectal cancer (160,000 new cases and 60,000 deaths in the United States each year) are compelling public health concerns. Following the evolution of effective surgery for this disease since the 1960s, the focus has been on improving methods of detection and integrating them into effective screening programs.This was the first study to evaluate the effectiveness, in a setting of comprehensive medical examinations, of using the fecal occult blood test in conjunction with sigmoidoscopy, rather than sigmoidoscopy alone, to screen for colorectal cancer. Our end points were extent of compliance with fecal occult blood test and sigmoidoscopy, numbers of cancers detected, and mortality rate.From 1975 through 1979, a total of 21,756 patients (aged 40 and older) who presented at the Preventive Medicine Institute-Strang Clinic for routine medical examinations were enrolled by calendar period into study and control groups. Study patients were offered annually both rigid sigmoidoscopy examinations and fecal occult blood tests requiring two stool specimens per day for 3 days, while control patients were offered only annual sigmoidoscopy. The majority of fecal occult blood test cards were not rehydrated before assay. Patients with positive tests were referred for double-contrast barium enema and colonoscopy. Two distinct trials were carried out. Trial I was primarily a demonstration of feasibility of using the fecal occult blood test as a supplemental screening method. Of the 9277 participants, 7168 (77%) were assigned to the study group and offered the fecal occult blood test. In trial II, approximately half of the 12,479 patients were assigned to each group. Patients in both trials had follow-up through 1984.Compliance with the fecal occult blood test was initially high in both trials, but diminished such that only 56% of study patients in trial I and 20% of those in trial II returned for second tests. On the initial (prevalence) screen, a substantial number of early-stage cancers were detected by the fecal occult blood test, primarily in trial II. In trial II, survival probability was significantly greater (P < .001) in the study group than in the controls (70% versus 48%), and colorectal cancer mortality was lower (0.36 versus 0.63) with borderline significance (P = .053, one-sided).The screening of average-risk individuals (aged 50 and older) for colorectal cancer through use of the fecal occult blood test in conjunction with sigmoidoscopy can increase the likelihood of early detection of this disease. This practice, coupled with prompt diagnostic work-up following positive tests, will result in treatment of earlier stage cancers and increased survival after treatment.

From 1974 to 1981, 1598 patients with non-oat cell carcinoma of the lung were seen and treated. All were staged according to the AJC staging system. Of these, 706 patients had evidence of mediastinal lymph node metastases (N2). There were 151 patients (21%) who had complete, potentially curative resection of their primary tumor and all accessible mediastinal lymph nodes. The histologic type of tumor was adenocarcinoma in 94 patients, epidermoid carcinoma in 46 patients, and large-cell carcinoma in 11 patients. The extent of pulmonary resection consisted of a lobectomy in 119 patients, pneumonectomy in 26 patients, and wedge resection or segmentectomy in six patients. Almost all patients also received radiation therapy to the mediastinum. Clinical staging of the primary tumor and the mediastinum was based on the radiographic presentation of the chest and on bronchoscopy. Before treatment, 104 of 151 patients (69%) were believed to have had stage I (90 patients) or II (14 patients) disease, and 47 patients had stage III disease, of whom only 33 had evidence of mediastinal lymph node involvement. Excluding deaths from unrelated causes, the overall survival rate was 74% at 1 year, 43% at 3 years and 29% at 5 years. Survival in patients with clinical stage I or II disease treated by resection was favorable despite the presence of N2 nodes (50% at 3 years). Survival in obvious clinical N2 disease was poor (8% at 3 years). There was no difference in survival between patients with adenocarcinoma and those with epidermoid carcinoma. However, survival was poorer in patients with N2 nodes in the inferior mediastinum compared to those without lymph node involvement at that level.

From 1984 to 1991, 136 patients with histologically confirmed non-small cell lung cancer and stage IIIa (N2) disease received two to three cycles of MVP (mitomycin + vindesine or vinblastine + high-dose cisplatin) chemotherapy. All patients had clinical N2 disease, defined as bulky mediastinal lymph node metastases or multiple levels of lymph node involvement in the ipsilateral mediastinum or subcarinal space on chest roentgenograms, computed tomographic scans, or mediastinoscopy. The overall major response rate to chemotherapy was 77% (105/136). Thirteen patients had a complete response and 92 patients had a partial but major response (> 50%). The overall complete resection rate was 65% (89/136) with a complete resection rate of 78% (82/105) in patients with a major response to chemotherapy. There was no histologic evidence of tumor in the resected specimens of 19 patients. The overall survival was 28% at 3 years and 17% at 5 years (median, 19 months). For patients who had complete resection, the median survival was 27 months and the 3-year and 5-year survivals were 41% and 26%, respectively. There were seven treatment-related deaths, five of which were postoperative deaths. To date, 33 patients, all of whom had complete resection, have had no recurrence after treatment. These results demonstrate that (1) preoperative chemotherapy with MVP produces high response rates in stage IIIa (N2) disease, (2) high complete resection rates occur after response to chemotherapy, and (3) survival is longest in patients who have a complete resection after major response to chemotherapy.

We assessed the effect of surgery on survival from stage I non-small-cell lung cancer based on data collected in these screening programs. The majority of patients diagnosed in each program were treated by surgical resection, but 5 percent of the Sloan-Kettering group, 21 percent of the Hopkins group and 11 percent of the Mayo group failed to receive surgical treatment. Appoximately 70 percent of the stage I patients in each program who were treated surgically survived more than five years, but there were only two five-year survivors among those who did not have surgery. We conclude that patients with lung cancers detected in stage I by chest x-ray film and treated surgically have a good chance of remaining free of disease for many years. Those stage I lung cancers which are not resected progress and lead to death within five years. Therefore, every effort should be made to detect and treat lung cancer early in high-risk populations. We assessed the effect of surgery on survival from stage I non-small-cell lung cancer based on data collected in these screening programs. The majority of patients diagnosed in each program were treated by surgical resection, but 5 percent of the Sloan-Kettering group, 21 percent of the Hopkins group and 11 percent of the Mayo group failed to receive surgical treatment. Appoximately 70 percent of the stage I patients in each program who were treated surgically survived more than five years, but there were only two five-year survivors among those who did not have surgery. We conclude that patients with lung cancers detected in stage I by chest x-ray film and treated surgically have a good chance of remaining free of disease for many years. Those stage I lung cancers which are not resected progress and lead to death within five years. Therefore, every effort should be made to detect and treat lung cancer early in high-risk populations.

Nine hundred sixty-one patients underwent operations for the treatment of carcinoma of the long; 18 percent of these were 70 years of age or older. The effect of various factors, singly or in combination, on the incidence of postoperative complications was assessed. Variables included age, sex, cardiopulmonary status, cell type, stage of the disease, and type of procedure performed. The postoperative course was uneventful in 81 percent of the patients; 8 percent had minor complications, 9 percent major complications and 2 percent died. The majority of the complications were cardiorespiratory. High risk factors were identified to be old age, restricted cardiopulmonary reserve, and the need for pneumonectomy. The low incidence of mortality and major morbidity observed was attributed to careful preoperative evaluation, selection of the appropriate surgical procedure, and inclusion of routine preoperative physiotherapy in all patients. Nine hundred sixty-one patients underwent operations for the treatment of carcinoma of the long; 18 percent of these were 70 years of age or older. The effect of various factors, singly or in combination, on the incidence of postoperative complications was assessed. Variables included age, sex, cardiopulmonary status, cell type, stage of the disease, and type of procedure performed. The postoperative course was uneventful in 81 percent of the patients; 8 percent had minor complications, 9 percent major complications and 2 percent died. The majority of the complications were cardiorespiratory. High risk factors were identified to be old age, restricted cardiopulmonary reserve, and the need for pneumonectomy. The low incidence of mortality and major morbidity observed was attributed to careful preoperative evaluation, selection of the appropriate surgical procedure, and inclusion of routine preoperative physiotherapy in all patients.

Our controlled trial of screening for colorectal cancer has now been in progress for almost five years. Screening is accomplished by rigid sigmoidoscopy in control and study groups and, in addition, by fecal occult blood testing in the study group. Patients screened are men and women age 40 and older, mostly at average risk. Fecal occult blood testing is with Hemoccult slides with patients on a meat-free, high-bulk diet without hydration, and with a four-day storage interval between slide preparation and testing. Patients with positive slides undergo diagnostic investigation that includes both colonoscopy and double-contrast barium enema and, in some, an upper gastrointestinal series. Preliminary results to date include: patient baseline statistics and subgroup comparability, rate of positive slides of 1–4%, predictive value for neoplasia of 44–50%, false-positives of 0.5–2.1%, favorable Dukes' staging of cancers in the study group, and high patient compliance. Considerably more follow-up is needed in our study and control population, and issues such as mortality and cost need to be addressed. Additional time will be necessary to provide firm conclusions.

From 1974 to 1978, 998 patients with carcinoma of the lung were treated at our institution. All were staged according to the staging system of The American Joint Committee for Cancer Staging and End Results Reporting. Of these, 445 patients had clinically evident mediastinal lymph node metastases. The condition of 204 was considered inoperable at diagnosis, based on histology of oat cell carcinoma, known distant metastases, or malignant effusion. Surgical exploration was undertaken in the remaining 241 patients and they were treated by resection and/or intraoperative implantation of radioactive sources. Of this latter group, 161 (69%) had either palliative resection or implantation of radioactive sources. The median survival in this group of patients was 9 months. There were 80 patients who had complete, potentially curative resection of their primary tumor and all accessible mediastinal lymph nodes. The vast majority of these were also treated by postoperative external radiation therapy to the mediastinum. In this group, there were 25 epidermoid carcinomas, 44 adenocarcinomas, eight large cell carcinomas, and three oat cell carcinomas. The survival rate was 73% at 1 year, 57% at 2 years, and 49% at 3 years. Survival at 3 years was slightly better in glandular carcinoma (adenocarcinoma or bronchiolar) than in epidermoid carcinoma (56% versus 44%). We conclude that there is a select group of patients with mediastinal lymph node metastases who can be effectively treated by combined resection and radiation therapy, with prolonged survival.

Radiographs of 168 patients with non-small-cell lung cancer were reviewed. Following a negative initial examination, 102 tumors were detected during routine annual screening, while 66 were diagnosed during the interval between screenings. The cancers detected on routine yearly examination were smaller; the rate of resectability was higher, a larger number were Stage I, and survival was better. Within the routinely screened group, 65% had evidence of cancer on reviewing earlier radiographs; these patients tended to have earlier stages of cancer and a better survival rate.

This is a report of the initial (prevalence) screening for lung cancer in a population of 10.040 cigarette-smoking men 45 yr of age and older, recruited from metropolitan New York. All had posteroanterior and lateral chest roentgenograms, and approximately half the men, randomly chosen, also had sputum cytology (dual screen). Fifty-three confirmed lung cancers were found; 22 (40%) were AJCC Stage I, with 85% survival for 5 yr after resection, whereas only 2 of the remaining men with Stage II to III lung cancer lived longer than 3 yr. In the dual screen group, in which the 2 detection techniques could be compared, 6 Stage I lung cancers were detected by radiology alone, 7 by cytology alone, and only 1 by both techniques. All of the cases detected by cytology alone were squamous carcinomas, whereas two thirds of those detected by radiology alone were adenocarcinoma.

From 1973 to 1981, 75 patients with T1 N1 M0 and T2 N1 M0 disease had a complete, potentially curative resection with mediastinal lymph node dissection. Thirty-eight had adenocarcinomas, 36 epidermoid cancer and one large cell carcinoma. Surgical treatment consisted of lobectomy in 54, sleeve lobectomy in three, and pneumonectomy in 18. Two patients died postoperatively. Of 17 patients with T1 N1 disease, 14 had no further treatment and three received postoperative radiation and/or chemotherapy; the 5 year cumulative survival rate of these patients was 56%. There were 58 patients with T2 N1 disease. Forty-five had no further treatment and 13 received postoperative radiation and/or chemotherapy; the 5 year cumulative survival rate of these patients was 48%. The overall incidence of local and regional recurrence was low, and the brain was the most frequent site of recurrence. Factors influencing recurrence were histology and proximity of the tumor to hilum. The specific nodes involved, the number of nodes affected, and the extent of involvement within the nodes had no observed effect on survival. There was no observed improvement in survival with the use of adjuvants. However, survival was significantly poorer in patients with visceral pleural involvement.

Study of the prevalence rates of uterine cervical carcinoma in situ among women attending centres of Planned Parenthood of New York City, Inc., showed a small but statistically significant difference between the population choosing and using the diaphragm and the population choosing and using oral steroids for contraception. This can be attributed either to a decreased prevalence rate for women using the diaphragm or to an increased rate for women using oral steroids. The reason for the difference is not apparent from these data.

Improvement in computing machine reliability through redundancy is studied as a function of the level at which the redundancy is applied. The reliability achieved by redundancy of complete, independent machines is compared to that achieved by redundancy of smaller units.

The Mayo Lung Project (MLP) reported lung cancer incidence and mortality in a population offered chest radiographs and sputum cytologic screening examinations every 4 months and a population offered only the Mayo Clinic advice to undergo annual examination. No mortality benefit attributable to screening was observed after 6 years of observation and at least 1 year of follow-up.The authors describe a simulation study designed to estimate from Mayo data the parameters in a mathematical model of the natural history of lung cancer and to estimate the potential benefit associated with periodic screening of high-risk people starting at 45 years of age.It was found that the mean duration of Stage I non-small cell lung cancer is at least 4 years and that rates of Stage I detectability and curability are less than 25% and 35%, respectively.A trial of the magnitude, duration, and contamination of the MLP would have a less than 20% probability of showing significant benefit from screening; however, long-term annual screening might result in a modest decrease in lung cancer mortality, ranging from 0% to 13%. A greater benefit would accrue from improved detection and treatment.

One-hundred-sixty-nine lung cancers have occurred to date among 10,040 cigarette smoking men who participated in the New York Lung Cancer Detection Program. Almost 40% of the cases, 65, were still Stage I when their disease was diagnosed; 62 had thoracotomy and resection, and in 57, mediastinal node dissection confirmed that the mediastinum was free of metastases ("true pathologic" Stage I). Fifty-four of the 62 (87%) are still alive at this time, while only 15 of 104 (14%) of those with Stage II and III lung cancers are alive. Only two patients of the 62 in Stage I who were treated by resection died of lung cancer, both with T2 tumors. Two others are alive with metastases, one died postoperatively, and five died of other causes without evidence of lung cancer. The estimated probability of survival for true Stage I lung cancer is over 90% at five years, and close to 40% of all lung cancers can be detected in this favorable stage by present radiologic and cytologic screening techniques.

The Memorial Sloan-Kettering Lung Cancer Detection Program is described, and data from this study are reported and analyzed. The program enrolled 10,040 cigarette-smoking men who were screened for at least 5 years by annual chest radiographs; half the men randomly selected also had 4-monthly sputum cytologic examinations. Of the 354 lung cancers that developed in these men during the screening period and a 2-year post-screening period, nearly half were adenocarcinomas and nearly one third were epidermoid (or squamous) carcinomas. Lung cancers detected by screening were more likely to be early stage (53 per cent) and resectable (65 per cent) than were those discovered by other means (for example, symptoms) or in the post-screening period (20 per cent early stage, 32 per cent resectable). Sputum cytology was most effective in detecting squamous carcinomas early; chest radiographs were most effective for detecting adenocarcinoma. However, sputum cytology did not reduce lung cancer mortality among these men who were in a program of annual chest radiographic examinations. Overall 5-year survival for all of the men who developed lung cancer was 35 per cent, compared with the national average of 13 per cent.

The purpose of this study was to determine (in survivors of 5 years after resection of their lung cancer) whether age, sex, histologic condition, and age have any influence on furthering survival beyond 5 years.From 1973 to 1989, 686 patients were alive and well 5 years after complete resection of their lung cancers. Survival analysis was carried out with only deaths from lung cancer treated as deaths. Deaths from other causes were treated as withdrawals. Multivariate Cox regression was used to test the relationship of survival to age, sex, histologic condition, and stage.The population in this study had the following characteristics at the time of operation: The male/female ratio was 1.38:1, and the median age was 61 years. The histologic condition of their lung cancer was adenocarcinoma in 412 patients, squamous cell in 244 patients, large cell carcinoma in 29 patients, and small cell carcinoma in 1 patient. The stage of the disease was stage IA in 263 patients, IB in 261 patients, IIA in 12 patients, IIB in 68 patients, and IIIA in 82 patients. The extent of resection was a lobectomy or bilobectomy in 579 patients, pneumonectomy in 55 patients, and wedge resection or segmentectomy in 52 patients. A recurrence or a new lung primary occurrence was considered as failure to remain free of lung cancer. The median follow-up on all patients was 122 months from initial treatment. Of the 686 patients, 26 patients experienced the development of late recurrence and 36 new cancers, beyond 5 years. Overall survival for 5 additional years after a 5-year check point was 92.4%. Likewise, survival by nodal status was 93% for N0 tumors, 95% for N1 tumors, and 90% for N2 tumors. Survival by stage was 93% for stage I tumors and 91% for stage II or IIIA tumors.In patients with surgically treated lung cancer, neither age, sex, histologic condition, nor stage is a predictor of the risk of late recurrence or new lung cancer. The only prognostic factor appears to be the survival of the patient free of lung cancer for 5 years from the initial treatment, with a resultant favorable outlook to remain well for 10 or more years.

A mathematical model of the progression kinetics of lung cancer in a periodically screened population is proposed and data collected by the Memorial Sloan-Kettering Cancer Center in New York are used for parameter estimation. It is assumed that the development of adenocarcinoma of lung is a stochastic process with two stages, early and advanced, characterized by mean times, detection probabilities, and cure probabilities. Confidence regions of these parameters are estimated using a number of novel techniques. It is found, surprisingly, that the mean duration of the early stage is at least 4 years, the detectability less than .2, and the curability less than .5. These estimates imply that annual radiographic screening from age 45 to 80 might decrease mortality from adenocarcinoma of lung by something less than 20%.

The early lung cancer detection program in New York is described, and preliminary prevalence rate data are reported. Cigarette-smoking men over the age of 45, who are considered at high risk, are encouraged to enroll in the program and, to date, 6,612 have done so. These men receive PA and lateral chest x-rays annually, while a randomly selected sub-population of about half also have sputum cytology every four months. A controlled study of sputum cytology as an adjunct to the chest x-ray is now in progress. There were 15 cases of late lung cancer (pathologic stages II and III) detected at the initial examination, and 11 early cases (pathologic stages O and I). In the population of 3,387 men who had x-rays alone there were three early lung cancers detected; in the population of 3,225 men who had both examinations four early cancers were detected by x-ray and four by cytology. In addition, three cases of larynx cancer were detected by cytology. The early lung cancers detected by x-ray were peripheral and most were bronchiolar or adenocarcinoma; the early cancers detected by cytology were central epidermoid carcinoma of major bronchi. No early carcinomas were detected by both techniques. These findings suggest that a combination of chest x-rays and sputum cytology can be more effective than either one alone in detecting early lung cancer.

Approximately 8000 cigarette-smoking men over the age of 45 have entered into a lung cancer detection program in New York City. Cytologic examinations of sputum were carried out on 4000 subjects and lung cancer was found by this technique in nine men with normal chest x-rays. Seven had in situ or incipient invasive epidermoid carcinoma confined to the bronchus. These seven cases were studied by detailed histologic examinations of the bronchial tree in the resected specimens through sixth generation subscgmental bronchi. It was concluded that: 1) invasive epidermoid carcinoma arises from carcinoma in situ of bronchial surface epithelium or an extension of that neoplastic epithelium in bronchial glands; 2) the site of origin is a segmental bronchus in most instances; and 3) each carcinoma should be considered as unifocal in origin even though there is a continuing risk of another primary lung cancer. It seems unlikely that squamous metaplasia or basal hyperplasia is an essential step in carcinogenesis; rather, we believe that carcinoma may arise in bronchial epithelium without regard to the presence or absence of basal hyperplasia or squamous metaplasia, which should be considered nonspecific reactions to injury that may or may not accompany carcinogenesis.

Decision theory, which has been greatly developed in the last two decades and which is today widely applied to economic, political and military problems, appears useful for decision making in cancer management. As an example, the application of its principles and practice to the so-called “prophylactic neck dissection” is presented. This question, whether or not to carry out a radical neck dissection in a patient with oral cancer without palpable neck metastasis, has been the subject of controversy for more than 50 years, with nearly as many surgeons favoring it as rejecting it. Application of the decision theory, which requires only simple arithmetic, greatly clarifies the problem and shows that, if certain hypotheses are accepted, in most cases the best results can be expected after prophylactic neck dissection only in patients with primary cancers over 2 cm diameter. For clinical use, a “decision table” was developed, which without calculation permits the cancer therapist to find the optimal decision immediately.

This paper is concerned with protocols for sequential clinical trials which meet the classical statistical criteria of significance and power and which tend to assign the majority of patients to the superior method of treatment. The object is to test the null hypothesis that there is no difference in mean survival times associated with two treatments. It is assumed that survival times are exponential with parameters, death rates, which depend upon the treatment and that patients arrive sequentially over a period of time and are assigned immediately to a single treatment. The assignment rules utilize the data collected about patients already treated to choose the treatment for the next patient. The trials are terminated when the likelihood ratio crosses a given boundary. It is shown that the operating characteristic is insensitive to the assignment rule, and that the expected number of patients receiving the inferior treatment can be markedly reduced by the use of data-dependent rules.

The problem treated is that of predicting the reliability characteristics of a complex system from data on individual components. A general model for systems maintained over a period of time is proposed, based on the idea that every system failure is induced by a component failure and corrected by the replacement of a single component. Moreover, it is assumed that components are sometimes replaced even when the system is operating correctly, in order to prevent unscheduled interruptions in operation. The assumptions which define the general model cover a number of different preventive maintenance policies, among them the following: (a) Block Changes: All components of a given type are replaced simultaneously, at times determined by a renewal process. (b) Individual Component Replacement on the Basis of Age: If a component reaches some given age without failing, it is preventively replaced. (c) System Check-Outs: If a component is used only intermittently and it fails while it is not being used, it does not induce a system failure until it is called into use. At regular intervals, those components which have failed without inducing system failure are located and replaced. (d) Marginal Testing: At regular intervals, a test is conducted to locate those components which are still operating satisfactorily but which are expected to fail in the near future. All components located by this test are replaced. It is assumed that preventive removals are regeneration points and that the performance of a component may be described by a distribution function $F(x: y)$, the probability that a component is removed by time $x$, given that it enters the system at $y$, where $x$ and $y$ are both measured from the time of the last preventive removal. $F(x: y)$ is the sum of $A(x: y)$ and $B(x: y)$, where $A(x: y)$ is the probability that the component is preventively removed by $x$ and $B(x: y)$ is the probability that the component induces a system failure by $x$. The integral equations which determine the following measures of system performance from $F(x: y), A(x: y)$, and $B(x: y)$ are developed: (1) the expected number of failures in a given time interval (2) the expected number of preventive removals in a given time interval (3) the reliability function; i.e., the probability of no failure in a given interval following a given system age. Results from Renewal Theory and the Theory of Regenerative Stochastic Processes, developed by W. L. Smith, are applied to the problem of exploring the asymptotic behavior of these quantities. Conditions sufficient for maintenance policies a, b, c, and d to meet the assumptions of the general model are precisely formulated, and the analysis necessary to derive $F(x: y), A(x: y)$, and $B(x: y)$ is carried out for each policy.

Journal Article Survival with competing risks and masked causes of failures Get access BETTY J. FLEHTNGER, BETTY J. FLEHTNGER Mathematical Sciences Department, IBM Research Division, T. J. Watson Research Center, Yorktown HeightsNew York 10598, U.S.A.aisen@watson.ibm.comrsst305@uvm.haifa.ac.ilyashchi@watsonibm.com Search for other works by this author on: Oxford Academic Google Scholar BENJAMIN REISER, BENJAMIN REISER Mathematical Sciences Department, IBM Research Division, T. J. Watson Research Center, Yorktown HeightsNew York 10598, U.S.A.aisen@watson.ibm.comrsst305@uvm.haifa.ac.ilyashchi@watsonibm.com Search for other works by this author on: Oxford Academic Google Scholar EMMANUEL YASHCHIN EMMANUEL YASHCHIN Mathematical Sciences Department, IBM Research Division, T. J. Watson Research Center, Yorktown HeightsNew York 10598, U.S.A.aisen@watson.ibm.comrsst305@uvm.haifa.ac.ilyashchi@watsonibm.com Search for other works by this author on: Oxford Academic Google Scholar Biometrika, Volume 85, Issue 1, March 1998, Pages 151–164, https://doi.org/10.1093/biomet/85.1.151 Published: 01 March 1998 Article history Received: 01 April 1996 Revision received: 01 July 1997 Published: 01 March 1998

We report the results of a 10-year, time-trend, case-control study in which serum cholesterol level was determined at several points in time preceding the diagnosis of colon cancer in a population of individuals who sought general checkups at an ambulatory care screening facility. Each of the 69 patients with colon cancer (32 men and 37 women) was matched with a control patient who was randomly selected. At the time of diagnosis, the patients with colon cancer had significantly lower serum cholesterol values than control patients (5.56 ± 0.31 mmol/L [SEM] vs 6.47 ±0.34 mmol/L). This difference did not vary with sex or Dukes' stage of the cancer. The percent of matched pairs in which the cancer patient had a lower serum cholesterol level increased from 42% at 10 years prior to cancer diagnosis to 77% at diagnosis. The ratio of serum cholesterol at each period to the level at time of diagnosis demonstrated an average decline of 13% during the 10 years prior to diagnosis for case patients vs an average rise of 2% in the same period for control patients. We conclude that individuals in whom colorectal cancer develops share the same level of serum cholesterol as the general population initially, but during the 10 years preceding the cancer demonstrate a decline in serum cholesterol level that is opposite to the rising level seen with age in the general population. (<i>JAMA</i>. 1990;263:2083-2085)

From 1949 to 1972, 4732 patients with carcinomas of the lung were seen. Of these, 118 patients were found alive at 10 years from their initial diagnosis and treatment. This number represents only 2.5% of all lung cancers seen because many have been lost to follow-up within the first ten years, and their true status is not known. The authors attempted to look at the patient characteristics that might have influenced long-term survival. A case-matched study was also done of patients with carcinoma of the lung who lived less than 2 years and who had the same age, sex, histology, and year of diagnosis. In the long-term survivors, the median age was 57 years, 2 of 3 had epidermoid carcinoma, 86% were smokers, 76% were symptomatic, and 70% had one or both parents alive at age 70 years or older. There was no difference in these variables in the case-matched group. Of the 118 long-term survivors, 91 were men and 27 were women. Sixty-seven percent had Stage I disease, 13% Stage II, and 20% Stage III. None had distant metastasis at presentation. Of the 118 patients, 115 had surgery. Eighteen of the 115 surgically treated patients also had postoperative external radiation. Three patients with Stage III disease received radiation without surgery. None had recurrence of their cancer, 27 developed new cancers, and 70% of these were new lung cancers. Although the majority of the long-term survivors had Stage I cancers, a third had Stage II or III disease. No long-term survivor died of his original disease, suggesting that a 10-year survival is adequate proof of cure in carcinoma of the lung. Cancer 53:1405-1408, 1984.

<h2>Abstract</h2> This report describes the continuation of a study on the occurrence of early cervical neoplasia in relation to method of contraception used by women attending Planned Parenthood of New York City, Inc. An earlier paper indicated a small but significant excess in prevalence rates of carcinoma in situ for steroid users compared with diaphragm users. We now estimate incidence rates of carcinoma in situ and lesser precancerous cervical lesions in steroid, diaphragm, and IUD contraceptors who are initially proved free of cervical neoplasia. These data do not prove any statistically significant differences although the estimates exhibit variations not inconsistent with our previous findings.

This paper is concerned with the analysis of the reliability of complex systems in which components are used intermittently and which are maintained in operating condition by component replacement. The idea that a failed component causes system failure only when it is called into use is expressed mathematically. Based on component failure distributions and usage properties, the system reliability and expected time to system failure are derived as functions of system age for two different maintenance policies. With both policies, a component is replaced whenever it causes system failure. In the first, this is the only maintenance, while in the second, system check-outs are conducted at fixed intervals and all components which have failed without causing system failure are replaced. The two policies are compared and, for the second, the dependence of system reliability on the maintenance interval is determined.

A mathematical model of the progression kinetics of lung cancer was described and used to shed light on the natural history of adenocarcinoma and large cell carcinoma of lung from data collected in screening studies of male smokers by the Memorial Sloan-Kettering Cancer Center and The Johns Hopkins Medical Institutions. In both programs, estimates of the mean duration of early-stage adenocarcinoma or large cell carcinoma of lung ranged upward from 4 years, and estimates of the probability of detecting it in early-stage disease ranged downward from .16. The probabilities of curing early-stage disease through surgical treatment were found to be at most .52 and .13 in the New York and Baltimore studies, respectively. These estimates led to the conclusion that expected reduction in mortality from adenocarcinoma and large cell carcinoma of lung as due to annual x-ray screening from age 45 to 80 years is not greater than 18% in New York and 5% in Baltimore.

In [3], a general model for the reliability analysis of systems under various preventive maintenance policies is postulated and analyzed. The integral equations that determine the expected number of failures, the expected number of preventive removals, and the survival probability function are developed. In the present paper, a particular model of a system subject to marginal testing is considered and explicit values of these performance measures are obtained. Under a marginal testing policy, the system is maintained in operating condition by replacing all failed components as soon as they fail and, at regular intervals, conducting a test to locate those components which are still operating satisfactorily but which are expected to fail in the near future. All components located by this test are replaced. In this model, it is assumed that a component may be in any one of $n + 1$ states, $0, 1, \cdots, n$, and, during normal operation, these states constitute a continuous-parameter Markov process in which state $n$ is the failed state. When a component enters state $n$, it is immediately replaced by one in state 0. The marginal test detects the state and states $k, k + 1, \cdots, n - 1$ are considered marginal. The test is performed at fixed intervals, and, if a component is found in the marginal state, it is replaced by one in state 0. Since this model provides for transitions from any operative state to any other state, recovery from the marginal state to the good state is permitted, a characteristic which was not allowed in the model of [3]. In addition, a choice of the level at which the component is considered marginal is permitted. The loss of generality lies in the assumption that the process is Markovian. As in [3], it is assumed that there is no dependence between transitions in different component positions and that every system failure is corrected by the replacement of one component, so that the problem of determining system performance measures is reduced to the problem of determining the corresponding quantities for a single component position. In the analysis of this model, we shall first analyze the Markov process in the absence of marginal tests and determine the matrix of probabilities $H_{ij}(t)$ that a component is in state $j$ at $t$ given that it is in state $i$ at 0, the distribution function $F_i(t)$ of time to failure for a component initially in state $i$, and $\mathscr{B}_i(t)$, the expected number of failures in a component position by $t$, given that the component is in state $i$ at 0. It is then seen that, with the marginal test performed at intervals of length $T$ and replacement of marginal components, the states of the component just before successive test points constitute a discrete-parameter Markov chain. The matrix of transition probabilities $p_{ij}$ for this chain and the probability $P_i(r)$ that the component is found in state $i$ at the $r$th test point are determined. Then the expected number of preventive removals in a component position, $U_p(t)$, may be expressed as the sum of the probabilities of finding the component marginal over all test points before $t$ while the expected number of failures, $U_f(t)$, is expressible in terms of the $P_r's$ and the $\mathscr{B}_i's$. The asymptotic values of $U_p(t)$ and $U_f(t)$ are determined by the stationary properties of Markov chains. Finally, the Markovian nature of this model makes it relatively easy to express $R(t; x)$, the probability of no failure in a component position in an interval $t$ following system age $x$, in terms of the matrices discussed previously.

Probability functions are defined for use in reliability studies of equipments which are maintained over a long period of time through replacement of components. These are: lifetime distribution function, lifetime density function, probability of survival, hazard, expected number of replacements, and renewal rate. Theoretical results of renewal theory are adapted to reliability studies of complex systems.

In clinical trials comparing two treatments, one would often like to control the probability of erroneous decision while minimizing not the total sample size but the number of patients given the inferior treatment. To do this obviously requires that one use a datadependent allocation rule for the two treatments rather than the conventional equal sample size scheme, whether fixed or sequential. We show here how this may be done in the case of deciding which of two normally distributed treatment effects has the greater mean, when the variances are assumed to be equal and known. Similar methods can be used under other hypotheses on the underlying probability distributions, and will provide a considerable increase in flexibility in the design of sequential clinical trials.

In 448 patients with positive axillary lymph nodes who were treated with mastectomy at Memorial Sloan-Kettering Cancer Center, New York, from 1973 to 1978, estrogen receptor (ER) status was associated with survival. With a median follow-up of 75 months, significant differences were noted in the 288 postmenopausal patients; ER-positive patients had better six-year disease-free survival than ER-negative patients (60% vs 45%), as well as better overall survival. These differences were true in subgroups with one to three and four or more involved nodes. The addition of adjuvant systemic therapy had no significant effect on either ER-positive or ER-negative patients. The need for new imaginative systemic programs in the subgroup of ER-negative postmenopausal patients with breast cancer with positive nodes is apparent.

An estimated 135,000 men and women developed lung cancer in the United States last year1Silverberg E Cancer Statistics, 1983 (based on rates from NCI SEER Program 1973–1979) CA-A Cancer Journal for Clinicians.Am Cancer Soc. 1983; 33: 16-22Google Scholar and 90 percent or more will die of that disease,1Silverberg E Cancer Statistics, 1983 (based on rates from NCI SEER Program 1973–1979) CA-A Cancer Journal for Clinicians.Am Cancer Soc. 1983; 33: 16-22Google Scholar, 2National Cancer Institute. Cancer Patient Survival Report No. 5. Bethesda: National Cancer Institute, 1976 (DHEW Publication No. 77-992)Google Scholar most within two years. Among male cigarette smokers, deaths from lung cancer equal all other cancers combined.3Melamed MR Flehinger BJ Zaman MB Heelan RT Perchick WA Martini N Screening for early lung cancer.Chest. 1984; 86: 44-53Abstract Full Text Full Text PDF PubMed Scopus (472) Google Scholar In women, the incidence of lung cancer soon will exceed that of carcinoma of the breast, to become the primary cause of cancer deaths.4Greenwald P Polan AK Lung cancer deaths among women (letter).N Engl J Med. 1979; 301: 274Google ScholarAt this time, the chest x-ray examination and sputum cytology are the only means available for detection of lung cancer at an asymptomatic stage. The Memorial Sloan-Kettering study of a lung cancer screening program, reported by us in this issue (see page 44) was designed to determine whether cytology screening as an adjunct to the annual chest x-ray examination could facilitate detection and reduce the mortality from lung cancer. All 10,040 participants in that study, who were cigarette smokers at high risk of lung cancer, were offered annual chest x-ray examinations; half, chosen at random, also were asked to submit sputum for cytologic examinations every four months.Early lung cancer detected by cytology proved to be slow-growing epidermoid (squamous) carcinomas of major bronchi. The study showed that in a program of annual chest x-ray screening conscientiously carried out, these squamous carcinomas can be detected radiographically at a later time, but while still localized. Thus, the addition of cytologic examinations to the annual chest x-ray examination does not have an effect on overall mortality from lung cancer. In a single screening, however, cytology may be of value as an adjunct to the chest x-ray film since there are no subsequent examinations.The Memorial Sloan-Kettering study was not designed to evaluate the effect of an annual chest x-ray examination on mortality from lung cancer. All of the participants were offered annual chest x-ray examinations. Thus, the data from this study cannot be used to compare radiologic screening with no screening. However, we have established that a program of annual chest x-ray examinations, with or without cytology, will identify approximately 40 percent of all lung cancers (46 percent of nonsmall cell carcinomas) in AJCC stage 1. Furthermore, 70–80 percent of the men with localized (stage 1) lung cancer treated by resection do not die of that disease. There are only two reasonable explanations for these facts. Either the screening is saving lives of some men who would otherwise have had disease undetected until it progressed to an advanced, symptomatic and incurable stage, or these men could have lived for many years unaware that they had lung cancer. To advise against efforts to detect lung cancer early, ie, by screening asymptomatic high risk populations at least by annual chest x-ray films, is to assume the second explanation is true. Since symptomatic lung cancer is more than 90 percent fatal, a decision not to screen is equivalent to a decision not to treat for cure.We therefore recommend annual chest x-ray examinations for the detection of early lung cancer in asymptomatic subjects who are at high risk (eg, cigarette smokers over the age of 50 years). Sputum cytology should be included as an adjunct to the chest x-ray examination only if x-ray screening is to be performed on a single occasion or as infrequently as once every five years. Since it is difficult and expensive to select, recruit and motivate high-risk subjects to comply with a single site screening program, and since successful screening requires medical facilities for prompt investigation and resolution of suspicious findings, we recommend that screening for lung cancer in high-risk subjects be incorporated within a comprehensive health care system, ie, through the private practitioner, health maintenance organization or general medical clinic rather than as a single site screening program. Certainly the weight of evidence at this time supports the prudent medical practitioner who recommends regular screening of the asymptomatic person at high risk of lung cancer. An estimated 135,000 men and women developed lung cancer in the United States last year1Silverberg E Cancer Statistics, 1983 (based on rates from NCI SEER Program 1973–1979) CA-A Cancer Journal for Clinicians.Am Cancer Soc. 1983; 33: 16-22Google Scholar and 90 percent or more will die of that disease,1Silverberg E Cancer Statistics, 1983 (based on rates from NCI SEER Program 1973–1979) CA-A Cancer Journal for Clinicians.Am Cancer Soc. 1983; 33: 16-22Google Scholar, 2National Cancer Institute. Cancer Patient Survival Report No. 5. Bethesda: National Cancer Institute, 1976 (DHEW Publication No. 77-992)Google Scholar most within two years. Among male cigarette smokers, deaths from lung cancer equal all other cancers combined.3Melamed MR Flehinger BJ Zaman MB Heelan RT Perchick WA Martini N Screening for early lung cancer.Chest. 1984; 86: 44-53Abstract Full Text Full Text PDF PubMed Scopus (472) Google Scholar In women, the incidence of lung cancer soon will exceed that of carcinoma of the breast, to become the primary cause of cancer deaths.4Greenwald P Polan AK Lung cancer deaths among women (letter).N Engl J Med. 1979; 301: 274Google Scholar At this time, the chest x-ray examination and sputum cytology are the only means available for detection of lung cancer at an asymptomatic stage. The Memorial Sloan-Kettering study of a lung cancer screening program, reported by us in this issue (see page 44) was designed to determine whether cytology screening as an adjunct to the annual chest x-ray examination could facilitate detection and reduce the mortality from lung cancer. All 10,040 participants in that study, who were cigarette smokers at high risk of lung cancer, were offered annual chest x-ray examinations; half, chosen at random, also were asked to submit sputum for cytologic examinations every four months. Early lung cancer detected by cytology proved to be slow-growing epidermoid (squamous) carcinomas of major bronchi. The study showed that in a program of annual chest x-ray screening conscientiously carried out, these squamous carcinomas can be detected radiographically at a later time, but while still localized. Thus, the addition of cytologic examinations to the annual chest x-ray examination does not have an effect on overall mortality from lung cancer. In a single screening, however, cytology may be of value as an adjunct to the chest x-ray film since there are no subsequent examinations. The Memorial Sloan-Kettering study was not designed to evaluate the effect of an annual chest x-ray examination on mortality from lung cancer. All of the participants were offered annual chest x-ray examinations. Thus, the data from this study cannot be used to compare radiologic screening with no screening. However, we have established that a program of annual chest x-ray examinations, with or without cytology, will identify approximately 40 percent of all lung cancers (46 percent of nonsmall cell carcinomas) in AJCC stage 1. Furthermore, 70–80 percent of the men with localized (stage 1) lung cancer treated by resection do not die of that disease. There are only two reasonable explanations for these facts. Either the screening is saving lives of some men who would otherwise have had disease undetected until it progressed to an advanced, symptomatic and incurable stage, or these men could have lived for many years unaware that they had lung cancer. To advise against efforts to detect lung cancer early, ie, by screening asymptomatic high risk populations at least by annual chest x-ray films, is to assume the second explanation is true. Since symptomatic lung cancer is more than 90 percent fatal, a decision not to screen is equivalent to a decision not to treat for cure. We therefore recommend annual chest x-ray examinations for the detection of early lung cancer in asymptomatic subjects who are at high risk (eg, cigarette smokers over the age of 50 years). Sputum cytology should be included as an adjunct to the chest x-ray examination only if x-ray screening is to be performed on a single occasion or as infrequently as once every five years. Since it is difficult and expensive to select, recruit and motivate high-risk subjects to comply with a single site screening program, and since successful screening requires medical facilities for prompt investigation and resolution of suspicious findings, we recommend that screening for lung cancer in high-risk subjects be incorporated within a comprehensive health care system, ie, through the private practitioner, health maintenance organization or general medical clinic rather than as a single site screening program. Certainly the weight of evidence at this time supports the prudent medical practitioner who recommends regular screening of the asymptomatic person at high risk of lung cancer.

The nonrandom pattern of the spatial arrangement of chromosomes in the human metaphase complement has been resolved into separate properties by statistical analysis of matrices of ranks of interchromosomal distances averaged over sets of 25 or 50 metaphases. Assessment of the effect of a mitotic-arresting agent, Colcemid, on each of the properties of the pattern shows that those properties for which a spindle fiber dependence may be postulated are disordered by Colcemid, while the tendency for the acrocentric chromosomes to associate, attributable to their mutual participation in nucleolar organization, is not impaired by Colcemid. Furthermore, this analysis has revealed that chromosomes 21 and 22 display an exceptional propensity to be generally associative, &lt;i&gt;i.e., &lt;/i&gt;to be close to all chromosomes; that property is obliterated by Colcemid. Since a mitotic-arresting agent is routinely used in the preparation of cells for cytogenetic analysis, that property has hitherto not been recognized.

Annals of the New York Academy of SciencesVolume 161, Issue 2 p. 670-679 COMPUTER-AIDED DIFFERENTIAL DIAGNOSIS OF HEMATOLOGIC DISEASES* Martin Lipkin, Martin Lipkin Cornell University Medical College New York, N. Y.Search for more papers by this authorRalph L. Engle Jr., Ralph L. Engle Jr. Cornell University Medical College New York, N. Y.Search for more papers by this authorBetty J. Flehinger, Betty J. Flehinger IBM T. J. Watson Research Center Yorktown Heights, N. Y. Cornell University Medical College New York, N. Y.Search for more papers by this authorLouis J. Gerstman, Louis J. Gerstman New York University New York, N. Y.Search for more papers by this authorM. A. Atamer, M. A. Atamer New York University New York, N. Y.Search for more papers by this author Martin Lipkin, Martin Lipkin Cornell University Medical College New York, N. Y.Search for more papers by this authorRalph L. Engle Jr., Ralph L. Engle Jr. Cornell University Medical College New York, N. Y.Search for more papers by this authorBetty J. Flehinger, Betty J. Flehinger IBM T. J. Watson Research Center Yorktown Heights, N. Y. Cornell University Medical College New York, N. Y.Search for more papers by this authorLouis J. Gerstman, Louis J. Gerstman New York University New York, N. Y.Search for more papers by this authorM. A. Atamer, M. A. Atamer New York University New York, N. Y.Search for more papers by this author First published: September 1969 https://doi.org/10.1111/j.1749-6632.1969.tb34098.xCitations: 7 * This project was aided by grant AMO6857 from the National Institute of Arthritis and Metabolic Diseases, National Institutes of Health and contract PH 43-67-1337 with The National Cancer Institute of The National Institutes of Health. AboutPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat Citing Literature Volume161, Issue2The Use of Data Mechanization and Computers inClinical MedicineSeptember 1969Pages 670-679 RelatedInformation

Accuracy of chest film screening by technologists in the New York early lung cancer detection programBJ Flehinger, MR Melamed, RT Heelan, CM McGinnis, MB Zaman and N MartiniAudio Available | Share

Consider a system of k components that fails whenever there is a defect in at least one of the components. Due to cost and time constraints it is not feasible to learn exactly which components are defective. Instead, test procedures ascertain that the defective components belong to some subset of the k components. This phenomenon is termed masking. The authors describe a 2-stage procedure in which a sample of masked subsets is subjected to intensive failure analysis. This enables maximum-likelihood estimation of the defect probability of each individual component and leads to diagnosis of the defective components in future masked failures.

HEME, a computer program for diagnosis-oriented analysis of hematologic diseases, accepts as input information about a patient and provides as output an ordered list of suggested diagnoses, an analysis of the logic behind these diagnoses, and a list of tests relevant to these diagnoses and not yet performed. The decision algorithm is based on Bayes’ Theorem. Each disease in the system is individually analyzed, and the probability that the patient has the disease vs the probability that he does not is calculated. Bayesian methods of statistical inference are utilized in that the prior probabilities of the diseases and the probabilities of findings in given diseases were initially estimated from the judgment of experienced hematologists with the intention that they be modified automatically as data are accumulated. This program is intended for use in teaching hematology, as an aid to diagnosis, and as a means for studying the diagnostic process.

DUBROW, HILLIARD M.D.; MELAMED, MYRON R. M.D; FLEHINGER, BETTY J. PH.D.; KELISKY, RICHARD P. PH.D.; KOSS, LEOPOLD G. M.D. Author Information

Methods are presented for the statistical analysis of the patterns of association among metaphase chromosomes in the conventional cytogenetic preparation of human peripheral lymphocytes. For each cell midpoints of the p and q ends and the centromere of each chromosome were digitized, these coordinates were used to compute a 138 by 138 matrix of distances between points, and the distances were then transformed to ranks. A separation matrix, Δ, was then computed by averaging the matrices of ranks across the replicate cells for each individual subject. The Δ matrix was then subjected to a two-dimensional nonmetric multidimensional scaling analysis. Plots of the resulting two-dimensional coordinates of the chromosomes represent summary metaphases. Comparative analyses of patterns of association found in the separation matrices and in the plots were carried out. Consistent patterns of association between groups of chromosomes were obtained for four human subjects in both Δ matrices and plots. The larger chromosomes (1–12, and X) were peripheral while the smaller chromosomes (13–22, and Y) were central and clustered. The acrocentric chromosomes (13, 14, 15, 21, 22) were especially clustered. The q points of most chromosomes were more peripheral than either the p or the c points. The summary plot seems to be a more idealized representation of the expected characteristics of the intact metaphase cell than is any single metaphase of the set from which the summary was derived.

Abstract It may be economically advantageous to both consumer and producer to enter into an incentive type contract in which the price paid for a product depends upon the outcome of an acceptance test. If no test is performed, a product of some minimum quality is produced and the consumer suffers a considerable penalty. The outcome of the test will reflect, to some degree, the quality of the product. If the consumer agrees to pay a premium dependent upon the outcome, the producer will be motivated to invest money in improving his product and both parties may expect to profit from the arrangement. This paper constitutes an unconventional approach to acceptance testing, in which the test is viewed not as a procedure to obtain information but rather as motivation toward product improvement. Admissible strategies are defined and their properties are studied. An application to repairable systems is considered in detail.

This paper is concerned with the relationship between the occurrence of metastases and the size of primary cancers. We consider two probabilistic characterizations of this relationship. First is the distribution function of tumor sizes at the point of metastatic transition; second is the probability that detectable metastases are present when the cancer comes to medical attention. The equation relating these two functions is developed and conditions for their being identical are explored. Since the tumor size at the point of metastasis is not usually observable, estimation of the first distribution requires the use of the EM algorithm. Nonparametric methods of estimating both functions are explored, with attention to the fact that tumors often fail to be measured, particularly those that are known to be metastatic. The methods are applied to the estimation of primary tumor size at the point of distant metastasis in lung cancer (epidermoid and adenocarcinoma) and colorectal cancer and at the point of nodal metastasis in breast cancer. Monte Carlo experiments confirm that the bias inherent in the methodology is acceptably small.

Recent developments in screening, diagnosis and treatment of colon cancer could lead to a reduction in mortality from this disease. Removal of adenomas, identification of risk factors, appropriate application of accurate diagnostic tests, and aggressive anatomic-surgical resection of colon cancers may already be having a favourable impact. Screening of average-risk populations over the age of 50 also offers promise in the control of this important cancer. The disease is of sufficient magnitude to deserve detection at an early stage with better prospects of patient survival, since screening tests with moderate sensitivity and high specificity are available. Flexible sigmoidoscopy and faecal occult blood tests are sufficiently acceptable to be included in case-finding among patients who are in the health care system. The results of current controlled trials involving more than 300,000 individuals for evaluating the impact of screening on mortality from colon cancer are needed before this approach can be recommended for general public health screening of the population. Further research is required to develop better screening tests, improve patient and physician compliance, and answer more definitively critical questions on cost-effectiveness. Mathematical modelling using current and new data can be used to determine the effectiveness of screening in conjunction with recommendations for primary prevention.

This article considers the situation in which a system consists of k components and a defect in any component causes a system malfunction. When a system malfunction occurs, test procedures restrict the cause to some subset of the Ic components. When that subset consists of more than one component, this phenomenon is termed masking. Typically, masking introduces two types of problems. First, it is desirable to estimate the “diagnostic probability”—that is, the probability, given a specified malfunctioning subset, that each of the masked components is the defective one. Second, when a set of historical data contains masked information, one would like to use this information to estimate the defect probability of each individual component type. The article discusses these problems in detail and derives two-stage procedures for estimation and inference.

A screening program of 10,040 cigarette-smoking men over 45 years of age was undertaken in an attempt to achieve earlier diagnosis, thereby increasing the cure rate, of oat cell lung cancer. Of the 155 men who were found to have lung cancer, 27 (17%) had confirmed oat cell cancer. Only one case was diagnosed at the first examination. The other 26 cases (called incidence cancer) were diagnosed by subsequent examinations. In 24 of the 26 patients, the tumor was not found until it was advanced (Stage III), and of these patients, only one is alive at 21 months follow-up. Two tumors were diagnosed as oat cell carcinoma at an early stage (Stage I), and both patients are alive with no evidence of disease at seven and 24 months. The screening program used in this study did not succeed in detecting oat cell cancer at an early stage.

This article considers the situation in which a system consists of k components and a defect in any component causes a system malfunction. When a system malfunction occurs, test procedures restrict the cause to some subset of the Ic components. When that subset consists of more than one component, this phenomenon is termed masking. Typically, masking introduces two types of problems. First, it is desirable to estimate the “diagnostic probability”—that is, the probability, given a specified malfunctioning subset, that each of the masked components is the defective one. Second, when a set of historical data contains masked information, one would like to use this information to estimate the defect probability of each individual component type. The article discusses these problems in detail and derives two-stage procedures for estimation and inference.

Product tests are generally intended to serve two purposes, first to expose those mechanisms which are likely to prove troublesome so that they may be eliminated and second to provide estimat)es of the performance parameters of the product. In this paper, systems in which the failure rate is the significant performance parameter and in which failures constitute a Poisson process are considered. It is assumed that every system failure may be classified as “correctable” or “intrinsic” and that every correctable failure which occurs during test leads to elimination of the mechanisms which caused it. Optimal test plans are developed based on the purposes of the test and the economic factors of importance to the manufacturer.

Abstract It may be economically advantageous to both consumer and producer to enter into an incentive type contract in which the price paid for a product depends upon the outcome of an acceptance test. If no test is performed, a product of some minimum quality is produced and the consumer suffers a considerable penalty. The outcome of the test will reflect, to some degree, the quality of the product. If the consumer agrees to pay a premium dependent upon the outcome, the producer will be motivated to invest money in improving his product and both parties may expect to profit from the arrangement. This paper constitutes an unconventional approach to acceptance testing, in which the test is viewed not as a procedure to obtain information but rather as motivation toward product improvement. Admissible strategies are defined and their properties are studied. An application to repairable systems is considered in detail.

In this chapter, the authors summarize some of their accomplishments in mathematical modeling of the progression kinetics of cancer. These range from a simple general parametric model of the life history of a cancer in the setting of a periodic screen to a specific nonparametric characterization of size-dependent production of metastases. The authors consider the point at which metastases first become detectable by techniques standardly used in the medical community as the point of metastatic transition. It is generally thought that control of mortality from cancer depends strongly on identification and treatment of the disease early in its development, while tumors are small and localized. The simplest mathematical model of the natural history of cancer is based on the idea that a person is cancer-free at birth, makes a transition into a preclinical cancer state at some random age, followed by transitions into the clinical disease state, and finally to death.

A comprehensive listing of the articles and departments included in 1963.

To the Editor: Dr. Weiss participated in one of the first systematic studies of periodic screening as a potential means of controlling lung cancer, the Philadelphia Pulmonary Neoplasm Research project.1Boucot K.R. Cooper D.A. Weiss W. The Philadelphia pulmonary neoplasm research project: an interim report.Ann Intern Med. 1961; 54: 363-378Crossref Google Scholar His reports of that work have become a classic source of reference. We thank him for his kind comments on our own study. We agree with Dr. Weiss that rigorous statistical proof of the value of periodic radiographic screening is lacking. We would strongly support a study to obtain that proof if we believed it possible to deny chest x-ray films to a randomized, high-risk “control” population with their informed consent. In the absence of irrefutable proof, we note the following softer evidence of value in periodic screening: a shift to earlier stage disease at diagnosis, an increase in the proportion of cases that are resectable, and better survival in the screened compared to the general population. We certainly see no justification for a change from the present medical practice of recommending annual chest x-ray examinations for asymptomatic, high-risk individuals. A decision to do otherwise, as we have noted,2Melamed M.R. Flehinger B.J. Screening for lung cancer (editorial).Chest. 1984; 86: 2-3Abstract Full Text Full Text PDF PubMed Scopus (10) Google Scholar is a decision not to treat lung cancer for cure since the symptomatic patient is essentially incurable. The proportion of early-stage lung cancer detected in our population, the percentage of cases resectable, and the length of survival all are greater than in the three early studies referred to by Dr. Weiss. Operative mortality today for treatment of lung cancer by lobectomy at Memorial Hospital is 2 percent,3Nagasaki F. Flehinger B.J. Martini N. Complications of surgery in the treatment of carcinoma of the lung.Chest. 1982; 82: 1925-1929Abstract Full Text Full Text PDF Scopus (215) Google Scholar well below the 10–20 percent mortality of lung cancer surgery 20 years ago reported by Dr. Weiss.4Weiss W. Problems in screening for lung cancer. In: Miller AB, ed, Screening in cancer. A Report of U.I.C.C. International Workshop, Toronto, Canada. April 24-27, 1978Google Scholar Thus, we are seeing better radiographic detection and better treatment today than was the case 20 years ago. Patients who have had one lung cancer are at higher risk of another, as Dr. Weiss notes. In our population the risk of a new, second lung cancer following complete resection of a first is about fivefold the initial risk of lung cancer for the entire population. There were 12 men with multiple lung cancers among the 291 who have developed the disease at this writing. Three had two synchronous cancers. Nine of the 166 men who were treated by complete resection of their first cancer have since developed a second primary lung cancer (metachronous cancers). Based on a calculation of time at risk, we have an incidence rate of 16 per 1,000 person years for new primary second lung cancers after complete resection of the first. Five of the nine second primary cancers were resectable, and two of the five men treated by a second resection are alive and free of carcinoma six years and three years following resection of their second cancers. We find that very impressive. However, we agree with Dr. Weiss and with the consensus of the UICC International Workshop on screening in cancer,4Weiss W. Problems in screening for lung cancer. In: Miller AB, ed, Screening in cancer. A Report of U.I.C.C. International Workshop, Toronto, Canada. April 24-27, 1978Google Scholar in which he participated, that the basic index of screening efficiency should be population mortality, not case survival rate. We have reported the mortality rate from lung cancer in our population, but do not have a comparable population of unscreened, cigarette-smoking men with which they can be compared. There are other statistical approaches to solve this dilemma. The most promising, we believe, is through mathematical modeling based on the kind of precise, hard data available to us from the New York Lung Cancer Screening Program. We are presently engaged in deriving such a model from our data, and hope to use that model to show the effects of radiographic screening at various different frequencies, for populations that receive treatment of varying efficacy. It is our hope that we will have Dr. Weiss' critical comments on that effort when it is ready for formal presentation.

Carcinoma of the lung is the most common cause of death from cancer in the United States. In considering lung cancer for possible chemoprevention trials, we have analyzed the data collected by the collaborative NCI program on early lung cancer. The data indicate that at least 12 years of study of 80,000 people at risk for lung cancer (adult male cigarette smokers) would be required to establish a 25% reduction in squamous carcinoma of the lung. No intermediate markers of developing lung cancer are presently available to shorten the observation period. It is concluded that a study of the magnitude required is not feasible at the present time.

Melamed, Myron R.; Koss, Leopold G.; Flehinger, Betty J.; Kelisky, Richard P.; Dubrow, Hilliard Author Information

Journal Article Sequential treatment allocation in clinical trials Get access B. J. FLEHINGER, B. J. FLEHINGER 1IBM Research Center, Yorktown Heights and Cornell University Medical College Search for other works by this author on: Oxford Academic Google Scholar T. A. LOUIS T. A. LOUIS 2Columbia University Search for other works by this author on: Oxford Academic Google Scholar Biometrika, Volume 58, Issue 3, December 1971, Pages 419–426, https://doi.org/10.1093/biomet/58.3.419 Published: 01 December 1971

Product reliabilities are predicted from past failure rates and reasonable estimate of future failure rates. Algorithm is used to calculate probability that product will function correctly. Algorithm sums the probabilities of each survival pattern and number of permutations for that pattern, over all possible ways in which product can survive.