| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Department of Anesthesiology, Perioperative, and Pain Medicine (Dr. Memtsoudis), the Department of Surgery (Dr. Zellos), Division of Thoracic Surgery, and Department of Surgery (Drs. Patil and Rogers), Center for Surgery and Public Health, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA; and Mathematica Policy Research, Inc (Dr. Besculides), Cambridge, MA.
Correspondence to: Stavros G. Memtsoudis MD, PhD, Hospital for Special Surgery, Department of Anesthesiology, 535 East 70th St, New York, NY 10021; e-mail: memtsoudiss{at}hss.edu
Abstract
Background: Reports on the temporal evolution in lung resection are limited. To elucidate temporal changes in the demographics of lung resections, we analyzed nationally representative data that were collected for the National Hospital Discharge Survey from 1988 to 2002.
Methods: Data collected between 1988 and 2002 were analyzed. Patients with International Classification of Diseases, ninth revision, clinical modification, procedure codes for lung resection were included in the sample. Three 5-year time periods were created (1988 to 1992, 1993 to 1997, and 1998 to 2002) to simplify the temporal analysis. Changes in the prevalence of procedures, age, gender, race, length of care, mortality, disposition status, and distribution by hospital size were evaluated. Trends in procedure-related complications were analyzed.
Results: Between 1988 and 2002, a total of 512,758 lung resections were performed. Comparing the earliest to the most recent time period, we found increases in the average age (61.1 years [range, 1 to 89 years] vs 63.2 years [range, 1 to 91 years], respectively), in the proportion of patients who were female (40.1% vs 49.6%, respectively), and in the proportion of Medicare/Medicaid patients (43.8% vs 49%/4.7% vs 6.7%, respectively). Decreases in the average length of stay (12.9 days [range, 1 to 358 days] vs 9.1 days [range, 1 to 175 days], respectively) and in the proportion of patients discharged to their primary residence (86% vs 79.5%, respectively) were seen. The proportion of patients who had undergone lobectomies compared to other types of lung resection increased. Mortality rates were 5% vs 5.4%, respectively, while the frequency of complications decreased.
Conclusion: We identified temporal changes in lung resection surgery that may help in the construction of health-care policies to address the changing needs of and financial burdens on the health-care system.
Key Words: epidemiology • lung • surgery
Over the last century, the surgical resection of lung tissue has remained an invaluable intervention in the treatment of pulmonary malignancies, traumatic lung injury, and a variety of infectious pulmonary diseases.12345 A number of factors, such as technical innovations and a dynamic patient population, have subjected the field of thoracic surgery to constant evolution.6 Knowledge of these temporal changes, especially as they are related to acute care hospitalization, is crucial to health-care providers and administrators in order to appropriately construct health-care policies in addressing the changing needs of and financial burdens on the health-care system.
Reports on the temporal evolution of lung resection are rare, limited by small patient numbers, reflect selected populations (usually, single academic centers), and include patients with only specific indications and/or procedure types for lung resection (eg, segmentectomy, lobectomy, or pneumonectomy).789
In this study, we analyzed nationally representative data, collected for the National Hospital Discharge Survey (NHDS) from 1988 to 2002, to elucidate temporal changes in the demographics of segmentectomies, lobectomies, and pneumonectomies for all indications. We also examined changes in in-hospital mortality and complication frequencies during the 15-year study period.
Materials and Methods
The NHDS
The NHDS public access database was acquired from the Centers for Disease Control and Prevention (Atlanta, GA). The plan and operation of the NHDS has been published in detail previously.10 In brief, the NHDS includes medical information collected annually since 1965 by the National Center for Health Statistics in order to compile nationally representative data on inpatient utilization of short-stay hospitals. The hospital universe includes Medicare-participating, noninstitutional hospitals of various sizes, exclusive of military, Veterans Affairs, and federal facilities in the 50 states of the United States and the District of Columbia. Hospitals included in the survey are required to have an average length of stay of < 30 days to be considered a short-stay facility, or to be a general medical or surgical hospital, regardless of length of stay. Facilities are also required to have at least six beds for patient use. Periodic updates of the hospital universe are performed to account for changes.
To ensure accurate, nationally representative sampling, the NHDS uses a complex three-stage probability design. Information collected in the survey included diagnosis and procedure codes (ie, International Classification of Diseases, ninth revision, clinical modification [ICD-9-CM]), age, sex, race, principal expected source of payment (ie, insurance status), length of care, hospital size, and patient discharge status. Weighted data, with weights derived from census data by the NHDS, are provided to generate unbiased national estimates from the sample (1% of all hospital discharges in the United States). Prior to 1988, different survey methods were used for NHDS.10 To avoid any potential bias introduced by the redesign, we only used data collected from 1988 forward.
Patient Selection and Analysis
Data collected for each year between 1988 and 2002 were obtained, read into a statistical software program (SAS, version 9; SAS Institute; Cary, NC), and concatenated. Patients with procedure codes (ICD-9-CM) for segmentectomy (32.3), lobectomy (32.4), and pneumonectomy (32.5) were included in the sample (Table 1
). Patients under the age of 1 year were excluded. To simplify the analysis of temporal changes in lung resection procedures, three 5-year time periods were created (1988 to 1992, 1993 to 1997, and 1998 to 2002). Changes in the prevalence of procedures were examined by time period. Changes in age, gender, race, length of care, distribution of procedures by hospital size, and disposition status were evaluated across time periods. The frequencies of procedure-related complications over time were analyzed by determining cases that listed ICD-9-CM diagnosis codes specifying complications of surgical and medical care, including complications peculiar to certain specific procedures (996.X), complications affecting specific body systems not elsewhere specified (997.X), other complications of procedures (998.X), and complications of medical care not elsewhere specified (999.X), as described previously.11
|
Changes in mortality between time periods were also assessed, but these analyses must be interpreted with caution, as the weighted number of patients having a fatal outcome in each period was below 9,000. Research suggests that weighted estimates between 5,000 and 9,000 are potentially unstable.12 The same caution in the interpretation of data related to individual complication codes should be exercised where indicated. The significance of the changes over time was assessed first using 
2 tests and then using Z-scores for categorical variables, and with general linear models for continuous variables.
Results
General
Table 2
describes the characteristics of the overall sample population (last column) and changes in the prevalence of studied variables over time. The majority of patients in the total sample were male, were white, had been discharged to home after their surgery, and were insured by Medicare. The average age was 62.4 years (range, 1 to 91 years), and the average length of care was 10.8 days (range, 1 to 358 days). The majority of cases occurred in hospitals with a bed size of 200 to 499. The number of lobectomies surpassed the number of pneumonectomies and segmentectomies combined (Fig 1
).
|
|
Over time, fewer patients were discharged to their home residence and an increasing proportion was discharged to short-term and long-term care facilities. An increase in the proportion of Medicare and Medicaid patients was also found, most recently representing > 55% of all patients. The proportion of procedures performed in the smallest and the largest hospitals (< 99 and > 500 beds, respectively) increased compared to that performed in midsize hospitals between the first and the second time period. However, this trend was reversed in the most recent study period.
The average length of care dropped by nearly 30% from an average of 12.9 days in period 1 (1988 to 1992) to 9.1 days in period 3 (1998 to 2002). Additionally, the proportion of patients who had undergone lobectomies increased steadily over time (period 1, 60.1% of total number of procedures; period 2, 67.4% of total number of procedures; period 3, 74.4% of total number of procedures), while the proportions of pneumonectomies (period 1, 16.4% of total number of procedures; period 2, 11.1% of total number of procedures; period 3, 10.4% of total number of procedures) and segmentectomies (period 1, 23.5% of total number of procedures; period 2, 21.5% of total number of procedures; period 3, 15.7% of total number of procedures) decreased.
Table 3 lists the 10 most common primary ICD-9-CM codes for diagnoses associated with lung resection procedures from 1988 to 2002. Primary malignancies of the lung remained the most common indication for surgery over time. More recently, indications such as lung abscesses and HIV-related indications have made the top 10 list of indications.
|
|
The mortality rate decreased from 5.0% in the first time period (1988 to 1992) to 4.1% in the second period (1993 to 1997), but increased again to 5.4% in the third time period (1998 to 2002). Age and length of care among patients who died did not vary significantly between time periods. An increase in the proportion of male patients who died in the hospital after surgery was found from the first to the second time period (74.1% to 83.8%, respectively). The proportion of male patients who died decreased in the third time period to 77.7%, but remained higher than the proportion in the first time period.
Changes in race distribution among patients who died were seen over time but are difficult to interpret because of the increasing number of patients for whom race is not reported. Changes in the primary source of payment over time were also found. The proportion of Medicare recipients increased from the first to the second time period (58.5% to 69.2%, respectively) and decreased during the third time period (63.0%), but remained higher than in the first period. The mortality rate of privately insured patients had the opposite trend. The percentage of deceased Medicaid recipients increased in each time period. The proportion of fatalities increased over time in hospitals with < 99 beds and steadily decreased in those with 100 to 199 beds. While the proportion of fatalities decreased from the first to the second time period and increased thereafter in hospitals with 200 to 299 beds, the opposite observation was made for larger hospitals.
Discussion
This study analyzed nationally representative data of lung resections in the United States between 1988 and 2002. We identified a number of changes over time related to patient age, gender, race, disposition status, expected source of payment, hospital size, length of hospital stay, and mortality. Temporally, patients undergoing lung surgery were older, more likely to be female, had a shorter length of stay, were less likely to be discharged to their home residence, and had similar in-hospital mortality rates over the time periods of the study. The frequencies of procedure-related complications decreased over time. While the specific reasons for these changes cannot be explained by our analysis, these trends are important to inform public policy and health-care administration decisions, and are discussed below in the context of the current literature.
Lung Resection Types and Diagnoses
We identified a change in the overall number of lung resections performed between time periods, with an increase in the mid-1990s (ie, from 1993 to 1997) and a subsequent decrease in the most recent time period (ie, from 1998 to 2002) back to the level seen in the first period (1988 to 1992). Lobectomy was the most commonly performed procedure, accounting for 67.0% of all resections. The proportion of lobectomies compared to segmentectomies and pneumonectomies gradually increased over time. This finding may be partially explained by advances in the screening for lung cancer, which is the most common indication for lung resection in our study. Screening may allow for earlier diagnosis when the spread of disease is limited, potentially explaining the decrease in the rate of pneumonectomies performed in more recent years. Furthermore, a reduced risk of local recurrence with lobectomy rather than sublobar resections for early-stage (ie, stage I [T1N0M0]) small cell lung cancer could account for a decrease in the rate of segmentectomies.13 Changes in the rates of lobectomies were also found in a recent Canadian study14 that evaluated trends of lung resections for lung cancer. Over time, the most common indications for lung resection surgery remained malignancies of the lungs and bronchi. Interestingly, HIV-related indications have made the top 10 diagnosis list during the most recent time period studied. This finding may be explained by the increase in the number of patients surviving this disease long enough for pulmonary manifestations (ie, malignant or infectious diseases needing either therapeutic or diagnostic intervention) to develop.15
Age, Insurance Status, Gender, and Race
A number of factors, including improved medical care, have led to an ever-increasing age among the general patient population. However, advanced age is not typically considered a contraindication for lung surgery,16 and the number of patients over the age of 65 years presenting for surgery has been steadily increasing.9 Our data are in concordance with this described phenomenon. Although the overall change in the average age has been relatively small (2.1 years), the impact of this change becomes obvious when viewing these data in conjunction with trends in the patients’ sources of payment. As the average age is approaching 65 years, the number of procedures paid for by Medicare also increased. Interestingly, the proportion of private payers increased as well, while the percentage of patients being paid for by other sources decreased. These results warrant further investigation in order to assess the financial impact on various aspects of the health-care system.
The proportion of women undergoing lung surgery increased substantially over the study period from 40.1% in the period from 1988 to 1992 to 49.6% in the period from 1998 to 2002. This change may be due to the marked decrease of the male/female incidence ratio of lung cancer and to the fact that women with local disease have been found to undergo surgery as an initial treatment more frequently than men.17
Race has also been implicated as a factor in undergoing surgical resection for the treatment of early-stage lung cancer. For instance, Bach et al18 reported a lower rate of surgical treatment among blacks compared to whites. Our data show decreases in the percentages of both whites and blacks undergoing procedures, while the percentage of patients with no race reported increased. Due to increases in the number of patients with race not reported, we cannot draw conclusions regarding this study category. However, an analysis of the NHDS participating hospitals found that facilities with high rates of missing data for race served populations with higher than average white/black patient ratios.19 Improved reporting and accuracy of race/ethnicity data will be vital to understanding and eliminating the potential disparities in the patients undergoing surgery for racial/ethnic groups.
Length of Care and Hospital Discharge Status
The average length of care after lung resection for all indications decreased dramatically over 15 years. This decrease may be explained by factors such as the identification of reasons for prolonged stays20 and ways to address them, the advent of new surgical techniques such as VATS,21 and the maturation of patient care pathways.22
We found that the number of patients being discharged from the hospital to short-term and long-term facilities rather than to home increased. Similarly, decreased lengths of hospital stays after cardiac surgery and increased rates of discharge to nursing homes and other skilled nursing facilities have been described in response to prospective payment and managed care in the 1990s.23
In-Hospital Mortality and Hospital Size
Lung resection surgery is associated with one of the highest perioperative mortality rates of any surgical procedure24 and has been reported to reach almost 70% for pneumonectomies performed after chest trauma.25 A number of factors have been identified to influence postoperative mortality including increasing age,26 extent of lung resection (pneumonectomy vs lobectomy),27 specialty of the surgeon performing the procedure,28 American Society of Anesthesiologists score of at least 3, increased operating time, and prolonged mechanical ventilation.29
Mortality rates changed little over time, yielding an average of 4.8%. The in-hospital mortality rates after lung resection for lung cancer that have been reported in cross-sectional studies varies.2728293031 Rates as low as 0.8%27 have been published and represent a marked improvement from 1971 when the in-hospital mortality rate was 13.5%.32 However, these selected reports may suffer from publication bias, as few centers report high mortality rates. Additionally, the relatively low mortality rates found in these studies may reflect the fact that data were collected at highly specialized institutions and only from patients undergoing resections for lung cancer. In contrast, our study used nationwide data from hospitals of all sizes and included patients undergoing resections for all indications, and hence is more generalizable and representative.
We compared the characteristics of the patients undergoing lung resections who died during their hospitalization to those who survived. Interestingly, those who died after surgery were on average 5 years older (67.1% vs 62.4%, respectively), were more likely to be men (78.4% vs 53.3%, respectively) and Medicare recipients (63.4% vs 47.7%, respectively), and had a longer length of care (15.2 vs 10.8 days, respectively) when compared to the overall sample.
Studies3334 have suggested that there are better outcomes if lung surgery is performed in high-volume hospitals. While a trend to increasing mortality over time in hospitals with < 99 beds was seen compared to a decrease in mortality in hospitals with a size of 100 to 199 beds, these data are to be interpreted with caution as they may potentially be unstable due to the small weighted sample size, as discussed previously. In contrast, the ratio of fatalities to total procedures performed seemed to be more favorable in hospitals with < 99 beds when compared to hospitals with 200 to 299 beds (3.3%/5.3% vs 31.5%/24.1%, respectively). While the complexity of cases at larger hospitals may account for this finding, our data do not allow a definitive interpretation.
In addition, male gender and increasing age have been implicated previously as risk factors for mortality after lung surgery.2631 However, the lack of important information, such as American Society of Anesthesiologists score or disease stage, in the NHDS prohibits a meaningful regression analysis and the determination of whether these factors are independently associated with mortality.
Procedure-Related Complications
Procedure-related complications were recorded in almost one quarter of patients. However, these numbers may be underestimating the true incidence of adverse events occurring after lung resection, due to the limitations of the NHDS database. For instance, patients who left the hospital after their procedure and then were readmitted for treatment of their complication are not captured. In addition, the entry of an event as related to the procedure may vary significantly by physician assessment and may be subject to individual coding practices.
We found a decrease in the frequency of procedure-related complications over time from 29.1% to 21.8%. This decrease was paralleled by a decrease in the prevalence of pulmonary complications. The reasons for these findings cannot be determined by our data. However, improvements in medical care and patient selection may be partially responsible for our findings.
Conclusion
This study provides valuable, nationally representative information on the evolution of lung resection surgery. Increasing age, the rise in the proportion of Medicare recipients undergoing lung resections, as well as increasing numbers of patients being discharged from the hospital to other health-care facilities have a significant financial impact on the health-care system. The rise in noncancer indications for lung surgery, like HIV-related diagnoses, is of importance to physicians, as they may need to expand their practice spectrum. The closure of the gender gap and the decrease in the length of care and in the frequency of complications are additional important findings of our study. Future research is needed to assess trends when controlling for potentially confounding factors. Our data may help in the construction of health-care policies to address the changing needs of and financial burdens on the health-care system.
|
Abbreviations: ICD-9-CM = International Classification of Diseases, ninth revision, clinical modification; NHDS = National Hospital Discharge Survey
The authors have reported to the ACCP that no significant conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article.
Received for publication April 1, 2006. Accepted for publication May 12, 2006.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
