Overview
DHCD does not perform onsite monitoring of EA program shelters by inspecting them at least once every three years. Without effectively monitoring these shelters, DHCD cannot ensure that pregnant women and families in the EA program consistently live in safe and sanitary conditions. For example, during our review of inspection reports, we noted that inspectors identified a variety of health and safety issues such as inoperable carbon monoxide and smoke detectors, electrical wiring issues, missing window guards, damaged doors and windows in need of repair or replacement, blocked entrances/exits to the shelter, and bedbugs. We also noted instances in which an inspector recommended that shelter rooms be taken out of service to be remodeled.
During our audit, there were two DHCD compliance coordinators (inspectors) responsible for inspecting the 929 shelter locations in the EA program. To comply with the department’s inspection practice, DHCD inspectors would have to inspect one-third of these 929 locations, or approximately 310 locations, each year. We reviewed all of the inspection reports for the 563 inspections DHCD conducted during our audit period. In fiscal year 2017, only 292 shelter locations were inspected. In fiscal year 2018, 271 locations were inspected; however, 81 of those locations had been inspected in 2017 and therefore no inspection was required in 2018. Additionally, none of the 563 inspection reports contained a DHCD supervisor’s signature. By not ensuring that all required inspections are performed, DHCD creates a higher-than-acceptable risk that health and safety issues may arise and go undetected and unresolved for prolonged periods.
Finally, DHCD does not collect the results of its annual reviews in one central location. Rather, annual inspections are stored on individual inspectors’ computers, and there is no procedure requiring them to centrally file these reports. As a result, DHCD cannot effectively monitor this annual review process to ensure that these reviews are conducted on time and that any identified problems are addressed in a timely and effective manner. Additionally, without historical data regarding issues identified during inspections, shelter contracts may be renewed when they should not be.
Authoritative Guidance
The contract between DHCD and the shelter contractors states that DHCD will monitor contract compliance through scheduled onsite inspections. Through discussions with DHCD management, we determined that DHCD has established a practice of visiting each shelter at least once every three years to ensure that the physical facility is well maintained, safe, and livable; the supervisor is then expected to review and sign each inspection.
Reasons for Issues
DHCD management stated that the department did not have sufficient staff members to perform these triennial inspections. In addition, it had no written policies and procedures for the performance of triennial inspections or the review of the inspection reports by a supervisor. DHCD also has not established a policy that requires inspectors to centrally file their inspection reports or a system that can be used to record and analyze them.
Recommendations
- DHCD should establish written policies and procedures that require a periodic (at least triennial) inspection of EA program shelters.
- DHCD should establish a process to collect, analyze, and store shelter inspection information in a central database and use this information to manage this process better.
- DHCD should establish a policy that requires supervisors to review and sign off on all inspection reports and submit them to DHCD to be entered in this central database before they are delivered to shelter contractors.
- Given the inefficiencies we identified in DHCD’s EA program shelter inspection process (e.g., inspecting shelters that were not due for inspection and not inspecting shelters that were overdue), DHCD should take the measures necessary to make this process as efficient as possible and then determine whether more staff resources are necessary to perform the inspections within the prescribed timeline.
Auditee’s Response
- The broad finding is inaccurate on its face, given that the audit report noted that 56 out of 60 shelters visited by the auditors were in "good condition" and only 4 were dirty, cluttered or in need of repair.
- The audit report incorrectly states the number of shelter units in the EA program. There are a total of 3,671 shelter units in DHCD's portfolio, not [929] as indicated in the audit report. It may be the case that the audit team was counting addresses at which shelter units are located, but some addresses contain multiple shelter units. Within the audit period, FY17 and FY18, DHCD inspected 3,106 shelter units, not 563 as indicated in the audit report. It may be the case that the audit team was counting inspection reports, but some reports reflect inspections of multiple units. In FY17, 829 congregate units and 632 scattered/co-shelter sites were inspected, for a total of 1,461 shelter units (40%). Similarly, in FY18, 723 congregate and 922 scattered/co-shelter sites were inspected, totaling 1,645 shelter units (45%). This rate of inspection puts DHCD well on schedule to inspect each shelter unit at least once every three years.
- The report states that inspections conducted during 2018 were not required with respect to 81 shelter units previously inspected during 2017. While it is true that units inspected in 2017 need not be re-inspected in order to meet the goal of inspecting each unit once every three years, when prior inspections are unsatisfactory, it is DHCD's practice to re-inspect the unit to confirm that health and safety issues were appropriately addressed. DHCD views this as an appropriate, targeted use of inspection resources and will continue to re-inspect units to ensure that issues found in previous inspections are addressed.
- As reported to the auditors but not acknowledged in the report, DHCD has been actively working on acquisition of technology to create a tracking system. . . .
- DHCD's current practice is to inspect at least 30% of our portfolio of shelter units annually. The department will memorialize in writing its existing policies and procedures that require a periodic inspection of EA program shelters. . . .
- Prior to the audit, [DHCD] had begun working to acquire the software necessary to create a centralized database so that shelter inspection information is collected and stored where it can be analyzed and used to better manage the inspection process. . . .
- Presently, all inspection reports are emailed by the inspector to the Contract Specialist who administers the contract with the shelter unit provider. All inspection reports are reviewed by the appropriate Contract Specialist. In addition, if an inspection report contains significant health and safety findings, the inspector emails the report directly to [the] Director of Contracts and Compliance, who oversees the inspection unit, to determine the appropriate action to be taken. When the new centralized database is created, DHCD will establish a policy that requires the inspection reports to be uploaded to the database with a provision for electronic sign-off by supervisors before delivery to shelter contractors. In the interim, the Director of Contracts and Compliance will sign all letters to shelter providers informing them of adverse inspection findings, and requesting the appropriate corrective action. . . .
DHCD agrees that it is important to make its inspection process as efficient as possible. DHCD believes that the inspection process observed by the auditors represents an appropriate triage of units. Inspection coordinators consider multiple factors in their efforts to maximize their time for efficiency, including the need to revisit units which may have ongoing issues. Inspection coordinators may also cluster inspections by region and site proximity. The audit did not identify any shelter units that were overdue for inspection; as explained above, the Department is on track to inspect every unit at least once every three years.
Auditor’s Reply
Our audit identified several problems with how DHCD monitors the conditions of its EA program shelters. Specifically, as noted above, it does not perform onsite monitoring of EA program shelters by inspecting them at least once every three years, its shelter monitoring process is inefficient, and it does not collect the results of its annual reviews in one central location so they can be effectively monitored.
Although DHCD is correct in pointing out that we only found problems at 4 of the 60 shelters we visited, this does not mitigate its responsibility to have an effective and efficient system in place to monitor its EA program shelters and promptly address any health or safety issues that may arise. Without effectively monitoring these shelters, DHCD cannot ensure that pregnant women and families in the EA program consistently live in safe and sanitary conditions.
During our audit period, DHCD did not have a process that accurately tracked all of its inspection activities; therefore, we cannot comment on the accuracy of DHCD’s assertion that it inspected 3,106 shelter units during this period. We found that DHCD only documented that it had conducted inspections at units where it identified problems and did not document that it had performed inspections at other units where it found no problems; thus there was no complete inspection record by program unit. During our audit, we asked DHCD’s Information Technology Unit to provide us with a complete list of EA program shelters by address. In response, the unit gave us a list of 982 shelter addresses, which we were able to verify as accurate and complete and which we therefore used to conduct our audit testing.
During our testing, we removed from our sample any reports that were made because of follow-up visits; therefore, none of the 81 shelter visits conducted during fiscal year 2018 at shelters that had been inspected the previous year were follow-up visits.
Based on its response, DHCD is taking measures to address our concerns in this area.
| Date published: | August 28, 2019 |
|---|