Brookings/ICF Long-Term Care Financing Model: Programmer's/Operator's Manual

U.S. Department of Health and Human Services

Brookings/ICF Long Term Care Financing Model: Programmer's/Operator's Manual

Peter Robertshaw and Lisa Maria B. Alecxih

February 1992

PDF Version (69 PDF pages)

This report was prepared under contract #HHS-100-94-0016 between the U.S. Department of Health and Human Services (HHS), Office of Disability, Aging and Long-Term Care Policy (DALTCP) and the Lewin Group. For additional information about the study, you may visit the DALTCP home page at http://aspe.hhs.gov/_/office_specific/daltcp.cfm or contact the ASPE Project Officer, John Drabek, at HHS/ASPE/DALTCP, Room 424E, H.H. Humphrey Building, 200 Independence Avenue, SW, Washington, DC 20201. His e-mail address is: John.Drabek@hhs.gov.

TABLE OF CONTENTS

OVERVIEW OF THE PROJECT

PREFACE

INTRODUCTION

I. MODEL DESCRIPTION

A. Overview

B. Code Structure

C. FORTRAN Subroutines

D. Assembler Subroutines

II. INPUT DATA SETS

III. OUTPUT DATAFILES

IV. RUNNING THE MODEL

A. Preparing to Run the Model

B. Running the Model

V. JOB CONTROL FILE

VI. UTILITY PROGRAMS

A. CHEKFL

B. TABLE2

VII. GENERATING A PERSON LEVEL FILE FOR TABULATIONS

ATTACHMENTS (separate file)

ATTACHMENT 1: DISLIFE.NEW -- Purchase probabilities

ATTACHMENT 2: DB.NEW -- Death benefits

ATTACHMENT 3: RPU.NEW -- Reduced paid up options

ATTACHMENT 4: PRE.NEW -- Monthly premium rates

ATTACHMENT 5: LAPSE.NEW -- Pension lapse rates

ATTACHMENT 6: Memo on Computer Budget for NIH Account

LIST OF FIGURES

FIGURE 1. Program Logic at the Subroutine Level

TABLE 1. Subroutine Descriptions

TABLE 2. Assembler Subroutine Descriptions

TABLE 3. Model Input Data Sets

TABLE 4. Summary of Model Output Files

TABLE 5. Example JCL for Model Execution

TABLE 6. Example DCL for Model Execution

TABLE 7. Sample Job Control File

TABLE 8. Sample Command Files for CHEKFL

TABLE 9. Input Parameters for the TABLE2 Utility

TABLE 10. Sample JCL and DCL for TABLE2

TABLE 11. Record Layout for Output Data File

OVERVIEW OF THE PROJECT

In September of 1988, the Office of the Assistant Secretary for Planning and Evaluation (ASPE) contracted with Lewin-ICF and the Brookings Institution to develop a public use version of the Brookings/ICF Long Term Care Financing Model. Using microsimulation techniques, the model projects the utilization and sources of financing for nursing home and home care services among the elderly to the year 2020.

Under this contract, many of the assumptions used in the model were revised to reflect data and findings that had recently become available. As the need for alternative policy simulations arose, the capabilities of the model were expanded. Examples of the types of simulations modeled include: the purchase of new private long term care insurance products; the use of pension funds to purchase long term care insurance; and publicly sponsored programs, such as the long term care benefits proposed by the Pepper Commission.

One of the products of this project is a public use version of the model code and accompanying documentation. The documentation includes:

Model Assumptions, which presents the assumptions used in developing the model.

Designing and Using Model Simulations, which presents assumptions used in modeling alternative proposals and using the results of the model.

A User's Guide to Specifying Simulations, which details how to specify simulations using the model's parameters.

A Programmer's/Operator's Manual, which shows the code structure and operation of the model.

PREFACE

This report is one of four related to the Brookings/ICF Long Term Care Financing Model. It outlines the programming logic, input and output data sets, and running the model. The three other documents discuss: 1) the assumptions used in developing the model; 2) assumptions used in modeling alternative proposals; and 3) using the results of the model, and how to specify simulations using the model's parameters.

This documentation was prepared by Peter Robertshaw, a subcontractor to Lewin-ICF and Lisa Maria B. Alecxih of Lewin-ICF. John Drabek, serving as the project officer, and Paul Gayer of the Office of the Assistant Secretary for Planning and Evaluation provided invaluable comments.

PROGRAMMERS/OPERATORS MANUAL

This document is intended to assist programmers/operators in the use of the Brookings/ICF Long Term Financing Model. Section I of this document is a description of the model. This description presents a brief overview of the model's purpose and a brief description of the programming logic at a subroutine level.

Section II and Section III of this document describe in detail the input and output data sets used by the model. Section IV discusses the editing, compiling, linking and running of the model from an operator's viewpoint. Sample JCL and DCL batch files are included to assist in the setup of the model. Section V gives details on the job control file (LTC2.CON) which is used to control important user-variable aspects of the simulation.

Section VI describes two utility programs for use with the model. The first is a table extraction utility, which gives the user the ability to extract a specified list of tables, suitable for printing or down-loading and importing into a PC-based spreadsheet. The second utility checks two parameter files and reports any differences between them.

Finally, Section VII describes a FORTRAN program that converts an optional data set produced by the model into a person level file from which tabulations may be produced.

I. MODEL DESCRIPTION

A. Overview

The Brookings/ICF Long Term Care (LTC) simulation model simulates the utilization and financing of both institutional and noninstitutional long-term care services for elderly individuals through the year 2020. Institutional services include nursing home care provided by skilled nursing facilities and intermediate care facilities. Noninstitutional services include home health, homemaker, choremaker, personal care, and meal preparation services. The model simulates the number of individuals receiving these services and the costs of these services as financed by various public and private sources. The overall objective of the model is to simulate the effects of various financing and organizational reform options on future public and private expenditures for nursing home and home care services.

The model begins with a nationally representative sample of the adult U.S. population. Each member of this sample population is assigned a weight value which, when aggregated, will approximate the total U.S. population in terms of both number of persons and demographic (i.e. age, sex, etc.) characteristics. Each record in the sample database contains information on a person's age, sex, marital status, income, assets, and other characteristics. The model simulates changes for each individual in the sample population from 1986 to 2020, simulating changes in age, economic status, disability status, utilization of long term care, and method of paying for such care.

The model uses a Monte Carlo simulation methodology. The model simulates changes in an individual's status by drawing a random number between zero and one and comparing it to the fixed probability of that event occurring for an individual with a given set of socio-demographic characteristics. In order to produce consistent results between runs, the random numbers used are drawn from a permanent datafile so that each time the model is run, the same random number is used for each decision made.

The model can be used to simulate long-term care financing assuming changes in private payment methods (such as increased purchase of private long-term care insurance), or new public financing programs. These simulations are greatly affected by the choice of assumptions about the economy (such as the rate of growth of the overall economy and nursing home prices) and individual behavior (such as rates of nursing home utilization and insurance purchase). The model can be used to make estimates using alternative assumptions to show how sensitive the results are to the assumptions chosen. Many of the possible assumptions and options are parameterized to allow for easy model operation.

B. Code Structure

The code structure for the Brookings/ICF Long Term Care Financing model is relatively straight forward. At its simplest level, the program consists of the main line program which calls a variety of subroutines to perform specific tasks. The main line program may be broken down into three sections (see Figure 1): initialization, data processing and reporting.

The initializing step sets up many of the arrays used in the model calculations, including defining tax rates, inflation factors, and poverty levels. Arrays used for accumulating and reporting functions are initialized to zero in this step.

The next step is the data processing loop. This loop is executed once for each record in the database. Intermediate arrays are initialized and then the annual cost loop is entered. This loop is performed for each individual for every year from 1980 to 2025. Within each year the routine checks for nursing home use, sets up random numbers used in the simulation, performs any 'special case' scenarios, checks for insurance availability, calls the costs calculation routines, stores costs in appropriate arrays, and calls the table accumulation routines.

The final step in the program is the reporting step. In this step, the main line programs calls a series of subroutines which total the accumulation arrays and print out the summary tables to the output file. This step is executed after processing the last record in the database.

C. FORTRAN Subroutines

The following tables describe the main line and each of the FORTRAN subroutines in the model. The description includes a listing of which routines call, or are called by, each subroutine. The main line program description is first, followed by each of the subroutines in alphabetical order.

D. Assembler Subroutines

In addition to the FORTRAN subroutines described above, the model also makes extensive use of three assembler language subroutines described in Table 2. IBM and VAX versions of these routines currently exist.

II. INPUT DATA SETS

Table 3 lists the input data sets required to run the model.

III. OUTPUT DATAFILES

Table 4 contains a listing and description of each output datafile the model creates.

IV. RUNNING THE MODEL

A. Preparing to Run the Model

The Brookings/ICF Long Term Financing Model is composed of 80 subroutines. In order to run the model, each subroutine must be compiled and linked into an executable module. This can either be done by 1) compiling each subroutine separately and then link-editing all the routines together, 2) or by combining all the subroutines into one file and compiling and linking that one file.

All the input datasets indicated as required [(1) = required dataset] in Table 2 must be available to the model when it begins running. In addition, any of the optional datasets called for in the parameter file must also be available. The output files from the model require approximately three megabytes of disk space, excluding the optional output datafile (see unit number 10 in Table 4). Approximately six megabytes of disk space are required to store both the input data bases and the output table file.

Finally, one tape drive is needed to read the input data base. See Attachment Six for a discussion of the cost of running the model at the National Institute on Health Computer Center.

B. Running the Model

Link-editing and executing this model requires approximately five megabytes of memory. Required run-time of the model depends on the type of computer being used. For example, cpu time on a VAX 8650 is approximately 95 minutes, and on an IBM 3090E cpu time is approximately 15 minutes. If the replacement population option is specified, these run times are approximately doubled.

Table 5 shows an example JCL batch file to execute the model on an IBM 3090E. Table 6 shows the equivalent DCL code to run the model on a VAX system.

V. JOB CONTROL FILE

The job control file is a small ASCII file that contains important parameters used to control the simulation. Table 7 shows an example job control file, and each of the 11 parameters in the file are described below.

VI. UTILITY PROGRAMS

Two utility programs, CHEKFL and TABLE2, are available to assist the user in the parameter setup and analysis of the model output. These utilities are described below.

A. CHEKFL

The CHEKFL program is a utility that allows the user to compare two input parameter files and report on any differences between them. The input parameter file is a fairly large file and when running multiple scenarios it sometimes may become difficult to check that only the desired parameters have been changed between two runs. CHEKFL can be used to compare the parameters for a new simulation against the base case or against a previous simulation scenario.

In addition, CHEKFL performs all logic, total, and data range checks on the input parameters that are required for a full model run. Sample JCL and DCL files to setup and run CHEKFL are shown in Table 8.

B. TABLE2

The TABLE2 utility program allows the user to define and extract a set of tables from the large table output file. The extracted file can be printed on the host computer, or download to a PC and printed locally. There is also an option to create a file that can be imported directly into a PC spreadsheet program.

A sample input control file for TABLE2 is shown in Table 9. The first line determines whether or not the optional Lotus readable output file is created (0=no, 1 =yes). The next lines are a list of the tables which are to be extracted (a maximum of 10 lines). The program reads in these lines and searches the table file for exact character string matches, and then copies the following records until the end of the current page is reached.

For example, entering TABLE 1 in the input file would cause the output of tables 1, 10, 11, 12, 13, ...,19. If you wanted table 1 only, you would enter TABLE 1# (the # identifier only works for tables 1-9). Table 10 shows example JCL and DCL files for compiling, linking, and executing the program.

VII. GENERATING A PERSON LEVEL FILE FOR TABULATIONS

The output dataset BASE.OUT may be converted into an ASCII/EBCIDIC person level file which can be used to produce additional tabulations. BASEOUT is written by the subroutine WRITR and is only produced when specified in the Job Control File through the parameter IWRITE. LTCCONV.FOR is a FORTRAN program that transforms arrays from BASE.OUT into a single record for every year after a person turns age 65 with demographic, economic, and long term care use information. The record layout for the converted database is shown in Table 11.

Brookings/ICF Long Term Care Financing Model: Programmer's/Operator's Manual

TABLE OF CONTENTS

ATTACHMENT 1: DISLIFE.NEW -- Purchase probabilities

TABLE 1.1. Purchase Probabilities: Disability Based Probability of Purchase

TABLE 1.2. Purchase Probabilities: Life Based Purchase Probabilities

ATTACHMENT 2: DB.NEW -- Death benefits

TABLE 2.1. Death Benefits: Individual Policies, 365-day Lifetime Maximum, and No Indexing

TABLE 2.2. Death Benefits: Individual Policies, 730-Day Lifetime Maximum, and No Indexing

TABLE 2.3. Death Benefits: Individual Policies, 1,825-Day Lifetime Maximum, and No Indexing

TABLE 2.4. Death Benefits: Individual Policies, 365-Day Lifetime Maximum, and 5%/Year Indexing

TABLE 2.5. Death Benefits: Individual Policies, 730-Day Lifetime Maximum, and 5%/Year Indexing

TABLE 2.6. Death Benefits: Individual Policies, 1,825-Day Lifetime Maximum, and 5%/Year Indexing

TABLE 2.7. Death Benefits: Employee Policies, 365-Day Lifetime Maximum, and No Indexing

TABLE 2.8. Death Benefits: Employee Policies, 730-Day Lifetime Maximum, and No Indexing

TABLE 2.9. Death Benefits: Employee Policies, 1,825-Day Lifetime Maximum, and No Indexing

TABLE 2.10. Death Benefits: Employee Policies, 365-Day Lifetime Maximum, and 5%/Year Indexing

TABLE 2.11. Death Benefits: Employee Policies, 730-Day Lifetime Maximum, and 5%/Year Indexing

TABLE 2.12. Death Benefits: Employee Policies, 1,825-Day Lifetime Maximum, and 5%/Year Indexing

ATTACHMENT 3: RPU.NEW -- Reduced paid up options

TABLE 3.A. % of Benefits Payable Under a Reduced Paid-Up Option by Duration: No Indexing

TABLE 3.B. % of Benefits Payable Under a Reduced Paid-Up Option by Duration: Indexing

ATTACHMENT 4: PRE.NEW -- Monthly premium rates

TABLE 1.A. Monthly Premium Rates: Individual Policies

TABLE 1.B. Monthly Premium Rates: Employee Policies

TABLE 1.C. Monthly Premium Rates: Retired Policies

TABLE 4.A. Monthly Premium Rates: Generic Policies

ATTACHMENT 5: LAPSE.NEW -- Pension lapse rates

Graduated Select Lapse Rates: Term Life Insurance

Graduated Select Lapse Rates: Pension Life Insurance

Graduated Select Lapse Rates: High Early-Cash-Value Insurance

ATTACHMENT 6: Memo on Computer Budget for NIH Account

ATTACHMENT 1. DISLIFE.NEW

Purchase probability tables. Probabilities by age and premium to income ratio, for disability based products and life insurance based products.

ATTACHMENT 2. DB.NEW

Death benefit tables by issue age, attained age and 12 options.