The HealthSoft
The RedAid Health Insurance tender presented a real challenge for HealthSoft, a software house
that specialized in hospital and pharmacy [Link] tender’s main item was an integrative
nationwide system for online handling of fees charged by hospitals for services, by pharmacies for
prescriptions, by physicians for clinic visits and by medical laboratories for tests. The tender also
included a comprehensive patient’s personal health information service to be made available
through the Internet. The customer’s Management Information System (MIS) Department was to
develop the home office modules, based on the existing software. In addition, the MIS Department
will purchase and install the hardware and communication equipment according to the contractor’s
specifications, see to the computerized interfacing agreements required with RedAid’s suppliers
of health services, and instruct RedAid personnel in the new system’s operation. All the systems
were to be under tight security, with high reliability requisite for all the components. The system
was to become fully operative not later than 13 months after signing the contract, with the
contractor fully responsible for the quality and timely completion of all system parts.
Already at the beginning of preparing the RedAid tender proposal, the HealthSoft tender team
realized that they needed the professional support of companies that specialize in software security
and data communication. The size of the anticipated programming load led the team to decide that
a subcontractor would carry out 60%–70% of the programming load. Cape-Code, a very small
software house located in a nearby suburb, was chosen as the programming subcontractor on the
basis of the lowest price proposed. Some “breathing space” when preparing the proposal was
obtained when the team discovered that the new enhanced Medal Software’s product Version 5E
of the widely used Medal Version 5, a laboratory accounting software program, included important
new modules. These new modules for online external authorization of patient credit and for the
preparation of monthly accounts for organizational customers like RedAid suited the tender
requirements. Medal’s developers had stressed the wide variety of their package’s interfacing
capabilities, which were touted as suited to almost any requirements. The
integration of Medal’s 5E version into the proposed software solved one of the remaining
difficulties hampering completion of the proposal and enabled substantial reduction of
development costs. Finally, HealthSoft signed agreements with all the potential external
participants – Lion Securities, Comcom and Cape-Code, subcontractors for security,
communication and programming, respectively – that framed its responsibility for financial issues
as well as coordination between the various organizations.
The day HealthSoft was announced winner of the tender was one of satisfaction and joy for the
company. Within a few days, all the project teams were working “at full speed”. Monthly
coordination meetings were conducted regularly. The subcontractors reported satisfactory progress
according to the project schedule. The first signs of alert appeared in the tenth meeting. Comcom,
the communication subcontractor, reported that some of RedAid’s major suppliers had refused to
supply the information needed for planning the communication equipment to be installed on their
premises as they had not reached an agreement with RedAid on the issue. As expected, Lion
Securities, the security subcontractor, faced similar difficulties. Both subcontractors declared that
even if full cooperation was to be achieved within a week, a one-month delay in completion of the
project was inevitable. Yet, Cape-Code people continued to express their satisfaction with the
progress of the development tasks they had undertaken. The next coordination meeting was a
special meeting, called after only two weeks, to discuss the severe delays that had appeared in
Cape-Code’s schedule. The delays had been discovered by a HealthSoft team when it tried to
coordinate a planned integration test. At this late stage, HealthSoft found out that Cape- Code had
subcontracted the development task to another small software house. It became clear that all the
previous calming reports had not been based on actual information; they were fabrications, meant
to satisfy HealthSoft people (and ensure regular income to Cape-Code).
Integration tests of the Cape-Code modules, begun 10 weeks behind schedule, identified many
more faults – of all kinds – than anticipated. Correction time required exceeded that planned. About
the same time, the team assigned to integrate the Medal Version 5E software into the system
realized that the enhanced version was not operative for all new modules, particularly the online
external authorization of patients’ credit status. In addition, the interfacing trials with other system
modules failed. Medal Software assigned a special team to complete the development of the
missing module parts and perform the necessary corrections. Though their efforts were visible,
successful completion of the software integration tests was accomplished almost 20 weeks behind
schedule.
The system test started 19 weeks behind schedule, with the same severity of quality problems that
had been observed at the integration phase. Finally, about five months late, it became possible to
install the hardware and software equipment in RedAid’s main office and at its suppliers’ sites.
The three-week conversion phase of the project, begun 23 weeks behind schedule, was,
surprisingly, a great success, with no major faults discovered and immediate repair of all faults
that were revealed. However, the implementation phase was a colossal failure: only one-third of
the staff listed for training actually participated in the instruction courses, and the majority of those
participating displayed insufficient preliminary knowledge of the new systems. Success with
supplier personnel was even lower. Only eight weeks later could regular operation of the system
begin, but with only about half of RedAid’s suppliers integrated into the new system.
The project, a frustrating one for all who participated, ended with a series of court claims. RedAid
sued HealthSoft, and HealthSoft sued RedAid, Cape- Code and Medal Software, the developers of
the Medal software package. Lion Securities and Comcom decided not to sue HealthSoft – despite
the extra costs they had incurred following RedAid’s lack of cooperation and the subsequent
obstacles raised to efficient performance of their parts in the project – in expectation of continuing
cooperation with HealthSoft on future projects. The trials lasted for years. The only consolation
was that the new software, once in operation, was a great success, with many of RedAid’s
management admitting that the system worked well beyond their expectations.
You may ask yourself:
■ Could the final gratifying results have been achieved without the “mess” experienced during the
course of the project?
■ Could they have been achieved without the major losses faced by all the participants?
■ Was the HealthSoft method of choosing subcontractors satisfactory?
■ Was the method of purchasing COTS software appropriate?
■ Was the method of controlling the implementation of the customer’s contribution
to the project adequate?
■ Was HealthSoft’s control over its external participants adequate?