compliance in smear positive tuberculosis patients?
Ahmet Levent KARASULU, Sedat ALTIN, Levent DALAR, Sinem Nedime SÖKÜCÜ, Pınar ÖZKAN
Yedikule Göğüs Hastalıkları ve Göğüs Cerrahisi Eğitim ve Araştırma Hastanesi, İstanbul.
ÖZET
Hastane yatışı yayma pozitif akciğer tüberkülozlu hastalarda kompliyansı artırır mı?
Çalışmanın amacı; akciğer tüberkülozunda hastaneye yatış ile tedavi başarısının ilişkisini araştırmaktır. Hastanede yat- mış insan immünyetmezlik virüsü (HIV) negatif, yayma pozitif 351 pulmoner tüberküloz olgusu çalışıldı. Tüm hastalar Dünya Sağlık Örgütü (DSÖ) kategori 1’e dahildi. Kronik ek hastalığı olan ve tedavide komplikasyon geliştiren hastalar çalışmadan çıkarıldı. Bölge dispanserinden takipli olan 306 hasta kontrol grubu olarak çalışmaya alındı. Gruplar tedavi- nin altıncı ayında tedaviye uyumları açısından değerlendirildi. Ortalama yaş 37.48 ± 13.87 idi; 204 (%31.1) hasta kadın- dı. Yatan hastalarda ortalama yatış süresi 25.4 ± 14.2 gündü. Toplamda 304 (%86.6) hastaneye yatan hasta tedavisini ta- mamlarken, geriye kalan 47 (%13.4) olgu takipten çıktı. Hastanede yatış süresi ile tedavinin tamamlanması arasında iliş- ki gözlenmedi (p> 0.05); 295 (%96.4) ayaktan hasta tedavisini başarıyla tamamladı. Tedaviyi tamamlamama olasılığı ya- tan hastalarda anlamlı olarak yüksekti (RR: 3.72 %95 GA: 1.96-7.05 p< 0.05). Sonuçlar göstermiştir ki; ek hastalığı olma- yan kategori 1 hastalarına tedavinin hastanede yatarak başlanmasının tedavi sonucunu olumsuz olarak etkilediği tedavi- yi tamamlayanların yüzdelerine bakılarak söylenebilir.
Anahtar Kelimeler: Tüberküloz, hastane yatışı, uyum.
SUMMARY
Can hospitalization provide better compliance in smear positive tuberculosis patients?
Ahmet Levent KARASULU, Sedat ALTIN, Levent DALAR, Sinem Nedime SÖKÜCÜ, Pınar ÖZKAN
Yedikule Chest Diseases and Chest Surgery Training and Research Hospital, Istanbul, Turkey.
Yazışma Adresi (Address for Correspondence):
Dr. Sinem Nedime SÖKÜCÜ, Ihlamur Sitesi C Blok Daire 5 Teşvikiye 80200 İSTANBUL - TURKEY
e-mail: sinemtimur@yahoo.com
Tuberculosis is as old as the history of mankind.
The history of medicine witnessed times when tuberculosis patients were isolated from the so- ciety; then came the age of sanatorium when rest and fresh air were considered sufficient for curing the disease; and finally the age of che- motherapy ensued (1). Ultimately we now all share the same contention that the disease sho- uld be managed on a social scale and global di- sease-control programmes are required (2).
World Health Organization (WHO) defined two tar- gets for the year 1991: diagnosis of 70% of smear positive cases and a cure rate of 85% (3-5).
The incidence of tuberculosis in Turkey lies so- mewhere between the incidence in communities with and without a successfully implemented tuberculosis control programme. In 2000, the reported incidence of tuberculosis in Turkey was 27/100.000 (6). This figure does not reflect all cases of tuberculosis, due to various problems in our reporting procedures. Some of those pati- ents who were diagnosed in governmental instu- titions or being followed in private practices are not included in this data. In 1980 and 1982, 25%
of the total population were considered infected (7). Each year 1-3% of these cases develop tu- berculosis (8). Therapeutic success rate in terms of cure and follow-up success is 80% (8).
On the other hand, a cure rate of 36.8% was re- ported due to the lack of bacteriological exami- nation (8).
In Turkey, 74% of the cases are initially hospita- lized (9). In Turkey, this general tendency toward
inpatient care continues to exist, as only 10% of the Turkish population is under DOTS administ- ration, and techniques for bacteriological diagno- sis have not been made widely available in local dispenseries (2). However, this seemingly app- ropriate approach results in serious difficulties as well. Prolongation of inpatient care, which itself is a costly service, causes a shift in the use of alre- ady limited resources in the direction of a stra- tegy with an uncertain benefit. Also, limited number of patient beds is another source of problem. A central management strategy emplo- yed in referral centers may lead to ignorance of local health services, thus delaying the solution.
Of course, if this approach can compensate for the drawbacks of the current control program and if it can provide better care for higher num- ber of patients, then potential problems can be afforded. Concrete data is needed to show whet- her this interim solution is beneficial. Our study was designed to answer this question.
MATERIALS and METHODS
A total of 2953 human immunodeficiency vırus (HIV) negative, smear positive cases of pulmo- nary tuberculosis who were treated in our hospi- tal from January 1997 to September 2001 were studied. All cases were WHO category 1 pati- ents (2). The patients in the control group, who attended to local dispensaries, were never hos- pitalized in any other hospital with the diagnosis or suspicion of tuberculosis before and during their antituberculous treatment other than a lo- cal dispensary for at least once.
The aim of the study was to explore the relation between hospitalization and the success of therapy in lung tuberculosis.
Three hundred and fifty one hospitalized human immunodeficiency virus (HIV) negative, smear positive cases of pulmo- nary tuberculosis were studied. All cases were World Health Organization (WHO) category 1 patients. The patients with chronic additional disease and those who developed complication during therapy were excluded. Three hundred and six outpatients attending to local dispensaries were also included as control group. The groups were compared with respect to completion rates at 6 months of therapy. The average age was 37.48 ± 13.87 years; 204 patients were women (31.1%). For inpatients, mean hospital stay was 25.4 ± 14.2 days. A total of 304 (86.6%) hospitalized patients completed their treatment;
the remaining 47 (13.4%) cases were not followed-up. No significant relation was observed between the length of hospital stay and completion of therapy (p> 0.05); 295 (96.4%) outpatients successfully completed their treatment. The probability of not completing the therapy was significantly higher for hospitalized patients (RR: 3.72 95% CI: 1.96-7.05 p< 0.05). Our results show that in category 1 patients without concomitant disorders, initiation of treatment at hospital has an adverse influence on the outcome of treatment, as reflected by the percentage of completers.
Key Words: Tuberculosis, hospitalization, compliance.
Following patient groups were excluded from the study: those who were followed up in another city than Istanbul (n= 1140); patients with unreliable medical records (n= 613), patients with diabetes mellitus (n= 228), chronic renal disease (n=
143), or chronic liver disease (n= 148); patients suffering from malnutrition (n= 92); homeless patients (n= 25); patients suffering from serious side effects due to medication (n= 193); patients who had been re-hospitalized due to serious ha- emoptysis during therapy (n= 20) [these are the patients for whom hospitalization was recom- mended in American Thorax Society/Centers for Disease Control and Prevention (ATS/CDC) gu- idelines]. For the remaining 351 cases, the data obtained from local dispenseries were checked against the data obtained from national center.
The completion status at 6 months therapy was stated.
A total of 306 age and sex matched outpatients attending to local dispensaries during the same period were also included as controls.
The 2 groups were compared with respect to completion rates at 6 months of therapy. The re- lationship between the length of hospitalisation and the completion of 6 months of therapy was explored.
Statistical Analysis
All data were analysed by SPSS 10.01(Statisti- cal Package for Social Science). Student t-test was used to compare the difference of length of hospitalization between the patients who comp- leted the treatment and those who could not be followed and mean age difference between two groups. The relationship between the length of hospitalization (numerical value) and completi- on rate at 6 months of therapy was evaluated by student-t test. Chi-square test was used to eva- luate the success of inpatient and outpatient tre- atment strategies. A p value less than 0.05 was considered statistically significant.
RESULTS
The average age was 37.48 ± 13.87 years (ran- ge 13 to 84 years); 204 (31.1%) patients were women, and 453 (68.9%) were men. The mean age of hospitalized patients was 38.4 ± 13.9 ye-
ars (range 13 to 84). The mean age of outpati- ents was 34.8 ± 13.9 (range 12 to 61). The two groups were similar with respect to age and gen- der distribution (student t-test p> 0.05).
For inpatients, mean hospital stay was 25.4 ± 14.2 days (range 6 to 89 days). A total of 304 (86.6%) hospitalized patients completed their treatment; the remaining 47 (13.4%) cases were not followed-up (Figure 1).
No significant relation was observed between the length of hospital stay and completion of therapy (25.04 ± 13.6 and 27.96 ± 17.5, p>
0.05) (Table 1); 295 (96.4%) outpatients suc- cessfully completed their treatment, while 11 (3.6%) cases could not be followed. The proba- bility of not completing the therapy was signifi- cantly higher for hospitalized patients (RR: 3.72 95% CI: 1.96-7.05 p< 0.05).
DISCUSSION
Although international health organizations such as WHO and IUATLD widely recommend and endorse implementation of DOTS with its pro- ven cost-efficacy and success, in Turkey, where there is a relatively long history of tuberculosis endemic, the treatment is initiated at the hospi- tal in 74% of patients regardless of referral sta- tus (9). Although chronic diseases, drug resis- tance, treatment failure, non-compliance to tre- atment, compromised health status, advanced disease or frequent episodes of haemoptysis,
Inpatient Outpatient 0
100 200 300 400
Completed Uncompleted
Count
Figure 1. Number of patients completed and un- completed treatment in both groups.
uncontrolled or insulin dependent diabetes mel- litus, chronic kidney or liver disease, drug al- lergy, adverse effects requiring hospitalization, coexistent illnesses that require hospitalization, conditions that pre-clude outpatient treatment, requirement for definite diagnosis, and being ho- meless were among the qualification criteria for hospitalization as stated by the central health authorities, this trend continues to exist.
On the other hand, as Ozkara et al. reported in their study, availability of DOTS and bacteriolo- gical diagnostic techniques at hospitals as well as the opportunity to recognise cases who are less likely to comply with treatment are among the assertions used to support this strategy (10).
The most important concern with regard to fol- lowing patients in their own social environment is the risk of conveying the disease to other members of family. The landmark Madras study and the subsequent studies have clearly shown that there is no difference with respect to infec- tion risk at home between outpatients inpatients and that the rate and percentage of improve- ment and the risk of recurrence are similar in these groups of patients (11,12).
Our study shows that in a large and densely po- pulated urban area like Istanbul with its remar- kable socio-cultural diversity and population dynamics, the completion rate was higher in outpatients with similar demographics compa- red to those patients who were hospitalized.
Given the fact that there is no comprehensive DOTS administration in our region, and patients get their full-month pill supply at one visit to a local dispensary and use them at home, and no bacteriological examination is performed at the end of the therapy, which is a prerequisite to de- termine the cure, it is clear that completer
analysis would be an insufficient measure. Des- pite this, the difference is striking.
Of the hospitalized patients, 13.4% were lost to follow-up, which represents a higher percentage compared to national average (9.2%) at the sa- me time period (9). On the other hand, the fact that the corresponding figure was only 3.6% in the outpatient group points out to the adverse influence of hospitalization on compliance to therapy. The causes of this adverse influence ne- ed to be examined in further studies. However, it does not necessarily mean that the patients who could not be followed, have not ever completed their treatment. It is possible that they might ha- ve completed their treatment without dispensary follow-up in somewhere which is out of national records such as in a private doctor office etc.
This issue might be a limitation of our study.
The longest hospital stay in our series was 89 days. The probability of achieving a negative culture result is 90% and the risk of recurrence is significantly reduced in a patient who has been hospitalised for 89 days, for whom DOTS should have been administered at least during the hos- pital stay. The total cost of the treatment regime consisting of 2 months of HRZS and 4 months of HR is 266 USD for outpatients, whereas the da- ily cost of inpatient treatment was 32 USD, as of August 2004, among 100 patients randomly se- lected from our hospital database, representing 2620 days of hospitalization. Since average length of stay in hospital was 25 days, the ave- rage cost is 800 USD only for hospitalization, and this value rises to 2560 USD for a 80-day stay in the hospital, reaching 10 fold of the cost of cure perpatient with DOTS. Thus, failure to fully implement one of these two strategies is the most important potential risk for Turkey and si- milar countries. These two strategies are either implementing a comprehensive DOTS control programme based on local units or diminishing the emphasis on local services by hospitalising all patients, therefore ignoring all economical di- sadvantages and inflicting an outdated manage- ment approach, as in the era of sanatorium. A dispute between the two approaches increases the risk of failure. A well-organised primary ca- re system is essential in our fight against tuber- Table 1. Mean hospitalization stay of patients
completed and uncompleted treatment in both groups.
Mean ± SD Range
Succesful 25.04 ± 13.6 7-87
Unsuccesful 27.96 ± 17.5 6-89
culosis. As shown in our study, substitution of primary health services in secondary centers and even in referral centers like ours inevitably results in a decline in our success rates; in other words, our quality of care decreases and the cost of treatment increases. On the other hand, DOTS is less expensive and more effective com- pared to traditional treatment strategies (1,13).
One of the contraversial points of our study is that we compared patients which are followed up only for 6 months of therapy. It is obvious that we cannot be sure that the patients who are not followed up or who moved to another city completed the therapy or not.
Actually this is the point that this study want to discuss and point to. In our countries control program there are breaks between layers of our health system. And this affects the tuberculosis control program unfavorably. From this aspect we believe all recorded cases must be compa- red. In the study why some patients were not fol- lowed up is clearly defined. On the other hand, fifth of the five elements of the DOTS is the co- hort analysis and this becomes impossible due to these breaks between layers (14).
One of the widespread ideas in our country is that the hospitalization in the inisial phase incre- ases the treatment coherence and by this way decreases development of MDR patients. On the other hand in ATS/CDC guides hospitalization was recommended only for non-adherant pati- ents. It was also, openly mentioned that fort he prevention of development of chronic patient the most important strategy is widen of DOTS. In the same guide the sources consumed for hos- pitalization can be spend for the education of the personel which takes part in DOTS and widesp- read DOTS.
Our results show that in category 1 patients wit- hout concomitant disorders, initiation of treat- ment at hospital has an adverse influence on the outcome of treatment, as reflected by the per- centage of completers. Thus, establishment of a nation-wide DOTS service is a strict requirement for an efficient tuberculosis control programme.
REFERENCES
1. Iseman MD. A Clinician’s Guide to Tuberculosis. New York: Lippincot Williams and Wilkins, 2002: 3-21.
2. World Health Organization. WHO Report 2004. Global Tuberculosis Control: Surveillance, Planning, Financing.
Geneva: World Health Organization, 2004.
3. World Health Organization. An Expanded Framework for Effective Tuberculosis Control. Geneva: World Health Or- ganization, 1991.
4. World Health Organization. Forty-fourth World Health Assembly, Resolutions and Decisions. Geneva: World Health Organisation, 1991.
5. World Health Organization. Stop Tuberculosis Initiative, Report by the Director General. Geneva: World Health Organisation, 2000.
6. World Health Organization. Global Tuberculosis Control.
Surveillance, Planning, Financing. Communicable dise- ases. Geneva: World Health Organization, 2002.
7. Bilgiç H. Tüberküloz epidemiyolojisi. Kocabaş A (editör).
Tüberküloz, Kliniği ve Kontrolü. Ankara: Emel Matbaası, 1991: 401-37.
8. Özkara S (editör). Türkiye’de Tüberkülozun Kontrolu İçin Başvuru Kitabı. Ankara: Rekmay Matbaası, 2003: 9-11.
9. Özkara Ş, Kılıçaslan Z, Öztürk F ve ark. Tuberculosis in Turkey with regional data. Toraks 2002; 3: 178-87.
10. Özkara S (editör). Türkiye’de Tüberkülozun Kontrolü İçin Başvuru Kitabı. Ankara: Rekmay Matbaası, 2003:
23-30.
11. A concurrent comparison of home and sanatorium treat- ment in South India Tuberculosis Chemotherapy Centre.
Bulletin of World Health Organisation 1959; 21: 51-144.
12. Toman K (editör). Tuberculosis, case-finding and che- motherapy. Questions and Answers. Geneva: World He- alth Organization, 1979: 122-9.
13. Weis SE, Foresman B, Matty KJ, et al. Treatment costs of directly observed therapy and traditional therapy for Mycobacterium tuberculosis: A comparative analysis.
Int J Tuberc Lung Dis 1999; 3: 976-84.
14. World Health Organization. Treatment of Tuberculosis: Gu- idelines for National Programmes, Geneva: WHO, 2003.
