You are on page 1of 9

Reference Laboratory Culture and DST registration and reporting system,

Combined Database
The system uses the following documents and computer files
1. Request and reporting form for culture examination of sputumV2, MS WORD
2. TB culture and DST laboratory performance report, Cult&DSTreportform.xls. It
also exists in a version that can call upon an automatic analysis program
CU&DSTreport_with_update_function.xls
3. Primary culture register, part of CultReg&DSTReg.xls. It also exists as a version in
which the left page can be filled automatically with information looked up in the
database, CultReg_with_automatic _fill.xls
4. Drug susceptibility testing (DST) and identification register, part of CultReg
&DSTReg.xls
5. Reference Laboratory analysis workbooks: CUanalysisbook.xls and
DSTanalysisbook.xls. They also exist as versions that can call upon automatic
analysis programs CUanalysisBook_with_automatic_fill.xls and
DSTanalysisBook_with_automatic_fill.xls.
6. Reference Laboratory database file, Reflab.rec: it must be installed together with its
associated files (.chk, .qes, .eix) in the same subdirectory. In case the name of this file
is changed when customizing the database and associated files for a specific country,
all the names of the files and references to these files in the automatic analysis
programmes under 7. here below (DSTrep.pgm and CUrep.pgm), will have to be
changed accordingly.
7. Reference Laboratory database analysis files: DSTrep.pgm with assisting files
DSTrep1.rpt, DSTrep2.rpt and DSTrep.txt for DST results analysis; and CUrep.pgm
with assisting files CUrep.rpt and CUrep.txt for culture results analysis. The name of
the laboratory-specific record file must be indicated in each of the .pgm files on the
first line. For instance READ c:\epidata\reflab\REFLAB.REC could be changed to
READ c:\epidata\reflab\UGREFLAB.REC in case UG has been added as the first
characters to the name of the database, for example for the country Uganda.
For automatic analysis including writing to the report forms above, the .rec, .pgm, .rpt, .txt
files above must be in the same subdirectory called c:\epidata\reflab\. This subdirectory
must also contain the Epi-Info 6.0 analysis.exe software (supplied with this set). The
correct name and path to the correct subdirectory must be specified in the .pgm files, if
placed elsewhere.
The output reports CU&DSTreport_with_update_function.xls, CUanalysisBook_with_
automatic_fill.xls and DSTanalysisBook_with_automatic_fill.xls can be anywhere.
The link to a .dbf version of the culture&DST database needed for the CultReg_with_
automatic _fill.xls will have to be adjusted to the subdirectory containing this Dbaseexported version, but they can be placed anywhere.
Description of the system and requirements
The forms and registers are recommended formats that can simply be printed and used
manually. More efficiently, they would be used partly manually, partly as computerized
databases with some associated computer-facilitated analyses, for internal monitoring of
culture performance, reporting of cultures and DST, redaction of working lists etc.. However,
these functions require basic knowledge of EXCEL and Epidata for initial customization and
proper usage. Knowledge of Epidata Analysis or Epi Info is not required to maintain the
1

database or to run these simple analyses, but Epi Info 6 Analysis must be installed to run
automatically in the background. For special manipulations using Epi Info may also be easier,
as explained under point 4 below.
For each specimen, a request form (1 above) has to be filled out by the person requesting
the culture and/or DST, and sent to the lab together with the specimen. The lower part of the
form can be used by the lab to return the results; this can be done in different steps (first
smear, later on culture and still later DST result), as shown on the form.
At the reference laboratory, a clerk enters the information from the request form in the
computer database (6). This database is the main document for all analyses, retrieval of strains
from the freezer collection, linking to patient data etc..., and for this reason it contains a large
variety of information. While entering the information from the request form, the clerk gives a
unique reference lab identification number to the specimen, enters it in the file and writes it on
the form together with the date of receipt at the lab, before the form is filed.
In contrast, the paper registers (3 and 4) are meant for bench work only, and for this reason
they only contain a small part of the same information that can be useful for the technicians at
the bench. Since many specimens do not require DST (not requested, or culture negative..)
and while the benches may be separated as well, there are two separate registers. The
specimen is preferably identified in both registers with the same unique ref. lab identification
number; if not, the DST register must contain a reference to the number of the specimen in the
culture register. These registers have several columns for registration of repeated and
quantified readings, results of identification tests . This will not be entered in the database,
but it is useful information to check when searching for the cause of a problem and should
thus be filled completely. Only the final result of all the different culture, identification or
DST tests performed on the same sample is recorded in the database.
The columns should be adapted to individual needs, depending on standard procedures:
o a reference column to the culture ID number in the DST register is not needed if one
and the same ID is used for all tests done on a sample
o the columns for drugs and their concentrations should be adapted to local practice, i.e.
types of drugs tested, one or more concentrations per drug, one or more bacilli
suspension strengths per drug, control suspension strengths. Also the identification
tests used should be individualized
The lines in the columns at the right of the register have been subdivided, to accommodate for
different media used for the same specimen (culture register) or for readings at different times
(DST register). These can of course also be adjusted to local needs.
The TB culture and DST laboratory performance report (2) is a quarterly report meant for the
NTP or any other authority which needs to be informed. It gives a summary report of both
culture and DST results batch-wise, i.e. for all tests set up during a certain quarter (or year).
In the automatic version, the figures needed for this report can be obtained from the database
by running the two analysis programmes (7), one for the culture part, and the other for the
DST part, by clicking on the knobs besides the tables and specifying the year and quarter or
month when asked to do so.
The two EXCEL analysis books (5) will compile the annual data from quarterly input, with
automatic calculation of some percentages. The cultures workbook will also turn the
percentages for internal monitoring into a graph (based on either quarterly, either monthly
counts). This graphical representation will make it easier to check for deviations from the

norm. In the automatic versions, the counts and filling is done automatically after clicking the
knob and specifying the month or quarter and/or year found besides the various tables in these
books. A monthly run is recommended for the cultures analysis, since this will provide the
figures needed for internal monitoring of culture quality.
Automatic filling of the left-hand pages of the culture register from the database can be done
only after exporting the .rec-file to an EXCEL or DBase file format under the same
subdirectory (export function of Epidata). At first installation, the link of the EXCEL sheet
formulas to the database will have to be re-directed. If the .rec-file was customized with added
or deleted fields, the formulas in the EXCEL register page cells may have to be adapted to the
changed sequence number of the columns. Filling the unique reference lab identification
number in the left column of the culture register, written (or better copied as a value)
exactly as it is in the EXCEL database-version will cause the other fields to be updated with
the appropriate information for the same specimen. This can be done for 1 page at a time, and
this L-hand page could then be printed to the verso of sheets with the R-hand page on the
recto side. From these the paper register can then be constituted, i.e. in a ring binder. From
these the paper register can then be constituted, i.e. in a ring binder. An example is included in
this set, REFLABTE.dbf and CultReg_with_automatic _fill.xls.
In a similar way, reports on individual sample test results can be made after customisation of
the database and of the reporting form format, using the EXCEL VLOOKUP function to add
the appropriate information to the form format for a specified unique identifier. In view of
individual preferences as well as database customization, no such automatic report form has
been included.
Daily working lists can be made directly from an EXCEL-exported extract of the database,
specifying the record numbers and fields wanted. It is not possible to select dates when
exporting a file. If desired, the output can also be adapted to include new columns to record
for instance the reference laboratory microscopy result, culture or direct DST inoculation
date, new identification numbers etc. After processing the samples and recording this
information on the lists, a clerk can then add these data to the database.
Other lists, for instance of pending results, can be made in the same way, using EXCEL to sort
the records in a way that will regroup those wanted. It is not possible to make such selections
in Epidata before exporting the file. However, this is relatively simple to do in Epidata or Epi
Info analysis, with writing of the file or list of fields wanted to a new .rec file, that can
eventually be exported to EXCEL for printing.
Since these are simple to make and in view of individual preferences as well as database
customization, no pre-determined format for such lists has been included.

Detailed use of the computer files


Database and associated files (6).
1. Overview
All files needed (.rec, .qes, .chk) have to be installed in the same subdirectory and have to
have the same name as the .rec file (exception for the extension). Data entry is meant to be
done using the software Epidata, and only this software will allow data entry with the two
rec-files related via the .chk files. In the combined file, data entry would also be possible in
Epi Info 6.04d after adapting the syntax of the .chk file to Epi Info.
The Epidata data- and automatic EXCEL analysis files provided ( Reflab.rec,
CUanalysisBook_with_automatic_fill.xls and DSTanalysisBook_with_automatic_fill.xls)
already contain data serving as an example. These can be overwritten during use, or the data
can be deleted using the Epidata software. Running the analysis EXCEL files on the empty
data files will then produce zero report and analysis sheets to start with. An easier way to
create a blank database may be to re-create one from the .qes file, at the same time re-naming
it after the country or reference laboratory. After giving new names to the databases, also
the .chk must be given the same name, and the database file name has to be changed in the
analysis programmes CUrep.pgm and DSTrep.pgm. The name of the laboratory-specific
record file always must be indicated in both the .pgm files on the first line. For instance
READ c:\epidata\reflab\REFLAB.REC could be changed to READ
c:\epidata\reflab\UGREFLAB.REC, for example for the country Uganda. The .eix will be
created with this same name automatically. Running the analyses in the
CUanalysisBook_with_automatic_fill.xls and DSTanalysisBook_with_automatic_fill.xls
on the new file is always possible if the new database is placed in the correct subdirectory
c:\epidata\reflab\, and will produce zero values to start with.
This user guide describes the combined database, containing patient as well as specimen
information. It is recommended if unique TB treatment numbers cannot be consistently used.
All information of the patient will have to be repeated for each of his/her specimens. It will
then be difficult to avoid confusion because of different spelling of names, making monitoring
of treatments or analyses by patient difficult.
2. Entering new records
o Using the combined file: first the lab identification number of the specimen will have
to be filled, or its components (in case these numbers consist of several parts, i.e.
study subject / year/ serial number) that then can be made automatically into a unique
identifier. The TBregnum field contains an autosearch function, a warning will appear
if the same number was already in the database. In case it concerns a new sample for
the same patient, click cancel when asked if you want to edit this existing record,
except if you want to check if the patient information is the same. The same
TBregnum can be used in several records, but once it exists in two of them, the
autosearch will no longer work.
o The file contains several fields to capture the origin of the specimen as administrative
divisions and finally the referring laboratory. It is strongly recommended to prepare
separate, simple record files under the same subdirectory, in which these data are
captured, so as to avoid seeing the same lab or division turning up several times during
analysis, because it was entered in different ways for various records. These file names
should then be added to the .chk file during country customization, as a Comment
legal ??????.rec SHOW command in the appropriate fields. It will then often also be

necessary to consistently change the designations of these administrative divisions in


the questionnaire as well as check file, wherever they occur. This will make the list of
available names or codes to pop up automatically, so that they can be entered by
clicking on the correct name.
3. Adding lab results
The lab results will have to be entered gradually later on, as they become available (first
microscopy, later again the culture result or molecular test results, and finally the DST result).
Individual records can be found by typing in the unique culture/DST reference lab ID number
in a new record, then answering Yes to the question if that record should be edited.
Alternatively, a search can be done on this (or another) field
- place the cursor in the field for the specimen or culture/DST lab identification number
- on the menu, click GOTO, then select Find record, or use CTRL_F
- it not yet shown, on the left side of the drop-down menu, choose the appropriate fields
using F3 to see the list.
- in the right column of the menu, type the search criteria, for instance the ID number
- confirm (click on OK), and the first record responding to the criteria will appear
- enter all new results for this record
- for further records, the procedure will have to be repeated entirely, unless nothing was
changed (in that case use F3 to go to the next record with same values for the search
criteria)
- Note: this serial search (F3) does not work with dates.
These search criteria are not kept in memory and are thus not practical for serial data entry.
This is better done defining a filter on the request fields designed for this purpose. These are:
Smear microscopy (Micr); Culture (Cult); Drug susceptibility 1st line drugs (DST1); Drug
susceptibility 2nd line drugs (DST2); Drug susceptibility genetic (DSTgen). Using these
filters needs the values in these fields to be updated consequently and consistently. This is
partly done automatically by the Epidata check-file as follows:
- microscopy request: if entered as not requested or unknown or impossible, the
field with Reference lab microscopy technique will be set to unknown, and the
Reference lab result to 9.0 (unknown). Otherwise, if requested is entered, this will
be changed automatically to completed after entering the result of the reference lab
- culture request: if entered as not requested or unknown or impossible, this will
automatically change DST first- and second-line to Not requested also; and the
result fields for solid and liquid culture will be set to not done. Once the field Date
culture inoculated is filled, this culture request field is automatically set to pending.
- request for DST on 1st respectively 2nd line drugs: if entered as not requested or
unknown or impossible, this will automatically set the DST results to unknown
(UUUUU appears in both the DST1 profile and DST2 profile fields). Once the
results of culture on solid / liquid media are filled, the DST1 request field is
automatically set to pending if culture was positive, but to impossible if negative,
contaminated, not done.... and the profiles are set to UUUUU. By default the DST2
is set pending only when a TB positive result of culture has been entered for a
retreatment case after Cat. 2 or MDR regimen, or for a MDR-contact case; in other
cases it will not be performed routinely together with the first-line DST.

request for genetic DST: if entered as not requested or unknown or impossible,


this will automatically set the results to unknown (U appears in both the RMP
genetic and Isoniazid genetic fields).
after entering a date in the fields for reporting of results (Date microscopy result
reported; Date culture result reported; Date DST result reported) the respective
fields for tests requested (microscopy; culture; DST 1st & 2nd line) will automatically
be set to reported

All these automatic entries can be changed manually, if needed. They will also be updated
after entering certain results:
- entering for both liquid and solid culture results OTHER THAN TBC will set the
culture request to completed, the DST request fields to impossible, and the DST1
and DST2 profile fields to UUUUU. Other entries in both the solid and liquid
culture results fields will set the culture request field to completed.
- after going through all fields for the DST results (various series), at the bottom of the
series a profile will appear. In this profile, U stands for unknown (not done,
contaminated..), - for susceptible, and the symbol of the drug for resistant, in the
following sequence: INH (H)-RMP (R)-EMB (E)-SM (S)-PZA (Z) -for first-line- or
Cm (C), Km (K), Ofl (O), PAS (P), Eto (E) -for the second-line drugs.
So, for instance the first-line drug profile HR--U means resistant to INH and RMP,
susceptible to EMB and SM, and PZA not done.
- with any entry in the 4 main first-line drug fields (INH, RMP, EMB and SM), the
DST1 request field will be set automatically to CPT or completed, except if this
DST1 field already contains IMP orNOT; the same happens for the request field
for 2nd line DST after entering the individual drug results (or U for unknown). Should
this CPT not be correct, the value can be changed back manually to RQU
(requested), allowing results to be added later on.
Adding results using a filter thus depends on consistent use of the test request fields, so that
the correct records will be retrieved. This is done by placing the cursor in the request field
concerned, selecting from the menu Filter, then Define filter, and entering in the box
RQU to show those with a requested test only. Only records with RQU in this field will
then be shown using F7 (previous) or F8 (next record), entering the new results and saving
while passing to the next record. After entering the results, the request field will be changed
automatically as described above. For culture or DST on first or second-line drug results, the
same system can be used defining a filter PND (pending) on the respective fields, provided
that the inoculation / setting-up dates were filled previously. If not, dates and results may be
added at the same time using a requested filter as above. After entering results, make sure
the profiles appear correctly, and the RQU in the request fields is changed to CPT. After
entering all new data, the filter must be deactivated (menu Filter, then Deactivate).
The same fields can be used to select complete results that need to be reported, setting the
filter to CPT. The value is automatically changed to RPT after entering the date of
reporting.
4. Other
Some operations can be better done in Epi Info
- revising the file structure after modifying the questionnaire file. To keep the
questionnaire and field names as they are, such a revision should be done using

Enter,Revise of Epi Info. If done in Epidata, curled brackets will appear in the
questionnaire screen around the parts of the field names.
sorting the file and re-writing it, for instance to facilitate entering large series of new
results. The REFLABTE file could then first be rewritten sorted on the specimen
identification, culture or DST number, as is easiest from the paper records that will be
used. Open the file in Epi Info Analysis, use the sort command and field name for the
sort, then rewrite the file under a new name (for instance, route
c\.....\REFLABT2.rec, then write recfile). Check if OK after opening the new
REFLABT2, then close and rename to the old name after erasing the original
REFLABTE.

Correcting duplicates in unique fields when the file refuses to open


If duplicate records are there, a warning indicating the record numbers of these duplicates
may appear and the file does not open. Write these down. If duplicate record numbers are not
given, but there is a warning about duplicates in a certain field, write down the field. Now
open the file in Epi Info 6, Analysis, and request a frequency on this field. This should show
which entries are there more than once. These duplicates can be corrected in Epi-Info
Analysis (update), or in Epidata after moving temporarily the corresponding .chk-file to
another directory, and activating it again by moving it back to the same (sub)directory as the
.rec file afterwards.
Correcting duplicates in unique fields when the file refuses to open indicating duplicates in
the eix file that are not real
It may happen that the files refuse to open, stating that there are duplicates, which are not
found. In this case, open the .chk file, inactivate the key unique from the commands under
the field that is the culprit by putting a star (*) in front of it, open the file, close it again and
remove the asterisk from the .chk. The files should now open again without problem.
How to interpret data entry in selected fields?
Field New or retreatment case in REFLABPA: this should make clear why the patient
sample is sent for culture / DST, eventually together with the next field Type of retreatment,
and give detailed information on the treatment regimen. New case should only be entered
for diagnostic specimens (before or very early in treatment), and not for contacts of MDR-TB
who should be entered as contact. All other patients have to be characterized by treatment
regimen, Cat.1, 2, 4 or 5, as explained by the drop-down menu (F9). The next field is only for
those patients (so remains empty for new or contacts), and must clarify the type of
retreatment case. Failure is then already used as soon as one month treatment has been
taken (for clarity, the duration on treatment can be entered in a field further on Month on
treatment, and another one Month of follow-up since end of treatment). For follow-up
samples during MDR-TB treatment or after cure, the original type of retreatment is kept
throughout, or if unknown U is entered. In the REFLABTE file the two fields mentioned
above are available to characterize only these follow-up specimens, by month during or after
treatment.

Analysis programmes (7)


Analysis of this database is best done in Epi-Info 6.04d Analysis mode, since the Analysis of
Epidata is still difficult to use for non-experts.
The system comes with 2 pre-programmed analysis tools (CUrep.pgm and DSTrep.pgm),
that can be activated also by people unfamiliar with Epi Info from the EXCEL reporting and
monitoring formats, i.e. CU&DSTreport_with_update_function.xls,
CUanalysisBook_with_automatic_fill.xls and DSTanalysisBook_with_automatic_fill.xls.
These will work properly only if all the necessary files have been copied to a
c:\epidata\reflab directory, with exactly these same file names.
- the report form CU&DSTreport_with_update_function.xls contains a part on
cultures as well as a second one on DST performed. The period is specified as a
quarter 3-6 months before the date of filling, but this can be changed to a month or
full year as required. To fill the data automatically from the database, two knobs on the
right of the form (Get&Update Culture resp. DST data) have to be activated
successively by a mouse click, after opening the file in EXCEL. This may require
adjusting the EXCEL settings to enable running of macros, needed to execute the
automatic reporting. The screen will first change to an Epi Info screen to allow
specification of the period to be analysed (if the screen does not change to Epi Info,
search a flickering Epi Info box on the lower Menu area, and click to open). When
prompted, enter the year to be analysed. A picklist now appears, in which the analysis
required should be indicated (by year, quarter or month). If quarter or month were
indicated, a second request will appear to specify the month or quarter. The program
will then run in the background, till the screen returns to EXCEL, with the question if
it should continue. Click on the OK or use Enter to make that the data are written
in the report cells. Printing the form will make it a permanent record, so that the file
can be re-used.
- the two books work in the same way, having knobs only for culture / only for DST
results, for each month and/or quarter. The year and month or quarter still have to be
selected in the same way as with the simple form. The months or quarters are added up
automatically to constitute annual data (in the sheet YR), with calculation of rates or
culture monitoring parameters, and updating of monthly or quarterly culture
performance charts (in the Chartmonthly or Chartquarterly sheets).
The counts for these reports have been defined as follows:
- the reporting period is defined by the date the specimen was inoculated for culture. So
this date is essential for the programme to include for counting all records it should
- DST or identification data are reported only for culture result (solid, liquid or both) TB
complex or other mycobacteria. Total counts are shown on the first line Total
inoculated. Incomplete first-line resistance profiles are not counted, except if it
concerns MDR (all H+R resistant considered). Completely missing first-line profiles
are reported as Pending, while those with one or more essential drugs not recorded
as Unknown are reported as Failed. Other mycobacteria (as entered in the culture
and/or species results) are reported as NTM. These are not counted for resistance
profile, except when it concerns mixtures, with somewhere a TB complex or TB result
entered, which should thus only be used for DST performed on a subcultured pure TB
strain. The totals taken into consideration for first-line drug profile appear on the line
Total M. tuberculosis complex.
o new cases and contacts are counted as new
o all defaulters are counted as RAD, return after default
8

relapses are counted as relapse


failures of Cat. 1 treatment are counted as failure
failures of Cat. 2 treatment, Cat. 4 and 5 patients are counted as Chronics
other or unknown (cases or retreatments) are counted as Other
second-line drug profiles are counted only for MDR-strains. Only the MDR
strains showing resistance also to one or more second-line drugs are shown
twice in the tables (once for MDR full profile, once as MDR + the type of
second-line drug resistance), but they are included only once in the totals for
the MDR
the culture report considers all records with any result in a culture field. The counts are
stratified by type of test, diagnostic (Suspects, not on treatment) versus follow-up
(Follow-up patients on treatment). For this re-grouping, all specimens of cases
entered as new, contact, relapse, return after default and other are counted as
Suspects if also nothing or zero is filled in the field for Month on treatment. All
others are counted as Follow-up.
o further stratification is by smear microscopy result; smears are classified as
positive, scanty, negative or unknown depending first of all on the peripheral
microscopy result; only when this is unknown is the reference lab reading
taken into account. For monthly or quarterly monitoring (workbook), positive
and scanty are grouped together.
o culture results are classified as positive, negative, contaminated and other,
based on one result per medium (so NOT per tube or bottle) that can be entered
in the database. Growth of mycobacteria (TB complex or NTM) on any
medium is considered as positive in this report. Positive on any medium
overrides negative or unknown on the other, while negative likewise overrides
contaminated or unknown. Unknown or missing cultures without any of the
other possible results on the other medium are counted as other. For monthly
or quarterly monitoring (workbook), only positive, negative and contaminated
results are considered.
o for the calculations and graphs, the denominator used for contamination is all
specimens, suspects as well as follow-up. For the other parameters, this
denominator is the total results (excluding other) only considering suspect
specimens. The percent false negatives takes into account negative cultures
from samples with positive or scanty smears.
o
o
o
o
o

You might also like